tko icon indicating copy to clipboard operation
tko copied to clipboard

This will ever become a version 4 of knockout?

Open nmocruz opened this issue 4 years ago • 8 comments

Sorry by maybe put some pressure and be a bit selfish by not respecting the good work and effort made but seems to dying before becoming a real version 4 of knockout. I fully understand the lack of time, I have also family and a job where the build of a framework is not so important as the client demands on features.

can something be done to help? how far is this to start a knockout version 4, or maybe tko beta?

nmocruz avatar Feb 08 '20 16:02 nmocruz

It does seem to have died:

https://github.com/knockout/tko/graphs/contributors

davidrwood avatar May 17 '21 14:05 davidrwood

yes, last year zero activity, I wonder if blazor will end like this, abandoned

nmocruz avatar May 17 '21 14:05 nmocruz

Not abandoned at all, still very actively used, but as it's being used in an early-stage company we're being extra careful about our support surface. Hopefully positive things to report in the near-future. 🤞

brianmhunt avatar May 18 '21 01:05 brianmhunt

near-future? sorry if it seems bold, but don't do like the aurelia team that promises something to next months and then stays 2 year releasing alpha versions. and that struggling to release a final version is very scary for us deciding what libraries to use. I belive that if was 2 years ago I was migrating my projects to ko 4 without douts but now seems like all other frameworks are much powerful and feature-rich then ko. I recently started to use ko-jsx to solve performance issues, seems not bad at all. I hope that tko can release be released, 3 year ago was already to late to some people that moved way.

nmocruz avatar May 18 '21 11:05 nmocruz

I've found this current release to be excellent. Much easier and more productive than React.

Martlark avatar May 20 '21 23:05 Martlark

@brianmhunt How large is the code base where it's being used? I know there's some work to do to move from 3.5x to TKO, but I am following with interest. We have a growing code code base, our ko components directory currently has around 5500 lines of code in html templates and 18,800 lines of JS in the view models for them (rough count in bash, nothing fancy). KO is still serving us well but keen to move to something more modern and I'm hoping TKO will be it :)

mattlacey avatar May 21 '21 01:05 mattlacey

@mattlacey Glad to hear your interest. The current largest code-base for TKO (that I know of) is around 100 KO web-components over 100,000 lines of JSX+JSS+Typescript.

I'm hoping to round out some comfort for TKO as a drop-in replacement, plus a bunch of docs that describe why it's worthwhile and how to get started. It's pretty close, but the big lingering item seems to be that TKO always uses its own CSP-safe parser data-bind, whereas knockout (without a plugin) uses new Function('...').

Interestingly, Svelte seems to be converging on many of the things TKO does, which is validating. They require a separate compiler though, whereas TKO uses standard & common JSX tools (which additionally makes bridging to/from React/Vue pretty trivial). Great to see interest in the front-end development, with projects like Svelte.

For the immediate moment I'm tackling the strangely unsolved build topology problem.

brianmhunt avatar May 28 '21 01:05 brianmhunt

JavaScript build systems are one of my least favourite things of all time, though JavaScript packaging comes close.

When the time comes I'd be more than happy to do some tests on a branch of our code base, to help gauge the level of effort involved and to help try and find any snags/hangups. I've not swtiched away from KO because it's not broken and works well for us, but it's always been on the back of my mind as something that's going to need to be handled at some stage.

On May 28, 2021, GitHub @.***> wrote:

@mattlacey https://github.com/mattlacey Glad to hear your interest. The current largest code-base for TKO (that I know of) is around 100 KO web-components over 100,000 lines of JSX+JSS https://cssinjs.org/?v=v10.6.0+Typescript.

I'm hoping to round out some comfort for TKO as a drop-in replacement, plus a bunch of docs that describe why it's worthwhile and how to get started. It's pretty close, but the big lingering item seems to be that TKO always uses its own CSP-safe parser data-bind, whereas knockout (without a plugin) uses new Function('...').

Interestingly, Svelte seems to be converging on many of the things TKO does, which is validating. They require a separate compiler though, whereas TKO uses standard & common JSX tools (which additionally makes bridging to/from React/Vue pretty trivial). Great to see interest in the front-end development, with projects like Svelte.

For the immediate moment I'm tackling the strangely unsolved build topology <https://github.com/knockout/tko/issues/134#issuecomment- 849287766> problem.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/knockout/tko/issues/108#issuecomment-850053181, or unsubscribe <https://github.com/notifications/unsubscribe- auth/AAJSWWYVP63BEFKRCBDMXITTP3ZCRANCNFSM4KR3RYBA>.

mattlacey avatar May 28 '21 01:05 mattlacey