Hackerskeyboard: Project dead?

[icebal]

@klausw
is this still being actively worked on? i've been a fan and have used your amazing board since it originally went onto the android store :) a release hasn't been made in a long while, and sadly there are no other alternatives out there that have a real layout. is it time to start working on the next Hackers Keyboard? maybe plans for a rewrite? if another maintainer is working on it, maybe a small not at the top of the readme? thank you again for the best, and only keyboard us system admins can use on that fateful page at 2am when we forgot to plug in the laptop all the way :)

[klausw]

Not entirely dead, but it is effectively in hibernation. I'd still like to get a new release out with some improvements merged, but realistically I'm not likely to be doing major changes at this point.

The underlying code is rather ancient, it's based on the Android "Gingerbread" keyboard from 2010, and a modern input method should do things like language selection in a very different way. I don't have the resources to test major changes, and I don't want to break things for current users.

I do think a rewrite as a separate project would be a good idea, especially if it could import existing layouts as a starting point. A fresh start would have the major advantage of not needing to worry about having to remain backwards compatible to existing configurations.

[icebal]

ive dabbled in programming when needed, but never a full project. as a newb
in android programming, would creating HK2.0 be a good first project?

On Wed, May 2, 2018, 12:51 PM Klaus Weidner [e-mail]
wrote:

Not entirely dead, but it is effectively in hibernation. I'd still like to
get a new release out with some improvements merged, but realistically I'm
not likely to be doing major changes at this point.

The underlying code is rather ancient, it's based on the Android
"Gingerbread" keyboard from 2010, and a modern input method should do
things like language selection in a very different way. I don't have the
resources to test major changes, and I don't want to break things for
current users.

I do think a rewrite as a separate project would be a good idea,
especially if it could import existing layouts as a starting point. A fresh
start would have the major advantage of not needing to worry about having
to remain backwards compatible to existing configurations.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
https://github.com/klausw/hackerskeyboard/issues/650#issuecomment-386063899,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AA5xlWpUgLeBqQbGK8E0j5tCyDroZs_Nks5tufIKgaJpZM4Tu1d_
.

[klausw]

I think this is going to be fairly challenging, I'm not aware of any modern and reasonably fully-featured open source keyboard to use as a starting point. In my case, I was able to start with basically just adding a new layout to the existing keyboard, all the extra options and features came later.

There's some background info here: https://developer.android.com/guide/topics/text/creating-input-method . But my impression is that the current IME sample code is rather basic, and it would need a lot of work to get expected functionality. For example, people had frequently been requesting swiping input, and I haven't seen an open source implementation for that.

[icebal]

awesome lol. i'll look around and see if i can find things new enough to
get a gist of what to do. all else fails, ill just start with the basic
android KB and go from there. thank you again for all your hard work on
this, it really has been the best keyboard hands down.

On Wed, May 2, 2018, 1:09 PM Klaus Weidner [e-mail] wrote:

I think this is going to be fairly challenging, I'm not aware of any
modern and reasonably fully-featured open source keyboard to use as a
starting point. In my case, I was able to start with basically just adding
a new layout to the existing keyboard, all the extra options and features
came later.

There's some background info here:
https://developer.android.com/guide/topics/text/creating-input-method .
But my impression is that the current IME sample code is rather basic, and
it would need a lot of work to get expected functionality. For example,
people had frequently been requesting swiping input, and I haven't seen an
open source implementation for that.

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
https://github.com/klausw/hackerskeyboard/issues/650#issuecomment-386069749,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AA5xlQcrvM_k_wlJ6382UD140wpxtXknks5tufZmgaJpZM4Tu1d_
.

[icebal]

also, you should think about having a "buy me a.beer/coffee" donate button somewhere. that way people can show their appreciation with something else other than empty words lol.

[jaysonesmith]

@klausw You open to merging and releasing more simple changes like #559 ?

[nift4]

AnySoftKeyboard got swipe.

[ZeChArtiahSaher]

Please, at least add option to import/export xml layouts in android. Having to decompile the apk just because I wish some buttons were on the front keyboard is a slow process. :(

[TPS]

@ZeChArtiahSaher Why _decompile_? It's fully open-source, so fork, add whatever from there, & open a PR to solve your #725, if you wish.

[klausw]

importing/exporting XML is not easy, at least not when using the existing XML resource files. There's some complicated inheritance going on for the language-dependent strings, and AFAIK the resource XML files get precompiled into a binary representation for inclusion in the APK file.

As far as "project dead" is concerned, there are two main issues. I have very little time to work on it, and I'm reluctant to add additional contributors with release privileges. A keyboard is privacy sensitive software, and I think it wouldn't be entirely fair to existing users to essentially hand it off to somebody else to run the project. The published APK gets pushed to all current users, whether they want it or not, and people get justifiably annoyed by regressions or just unexpected changes.

On top of that, the code rather fragile due to lack of tests and complicated internals (there's a reason why many applications resist adding hundreds of options to their software). The code is old and a bit creaky overall.

I'm all in favor if someone or a team wants to fork the project and make their own releases, or reuse existing layouts and other parts on top of a more modern codebase, but this should be a separate Play Store package where people would need to explicitly switch over if they want to use the new and more actively maintained version.

[LongJohn-Silver]

I guess it's safe to assume that it's dead now?

[jaysonesmith]

@LongJohn-Silver There was a PR merged 15 days ago, so potentially not?