Happy birthday + future/roadmap ?

[Erkan Yilmaz]

Happy birthday!

RT @myleneb To celebrate 2 years of #twister @mfreitas will you take a little time to tell us how you consider Twister's future ?
RT @kseistrup the zeroth block was mined 2013-11-14T01:57:35Z

@erkan_yilmaz

[Miguel Freitas]

Hi Erkan,

I believe most important landmarks for twister's roadmap have already been put...

I mean, to start, we obviously need to increase userbase. That won't happen if we stay mostly linux-based.

A good windows installer, that also manages running twister as a service when computer boots, is a first and easy candidate.

Then, a mobile client. We already cross-compile twisterd for android and ios, but packing it nicely with a good (native) UI is still missing.

The last nice client to foster usage would be the web-only client. We have already most of support implemented in twisterd (commands exposing raw data without crypto/authentication), a nodejs proxy prototype (to handle limiting traffic for some commands etc) and working javascript crypto. Still to decide: policy to allow torrent joining requests (to prevent server DoS) and packing client as a browser extension (to ensure code authenticity).

Ideas on improving user registration in general (fighting massive registrations) are important as well, but i don't currently see these as the most impeding factor.

These as just a few thoughts anyway, of course i might be overlooking a lot of cool ideas!

regards,

Miguel

[Cathal (Phone)]

Didn't someone make a Golang router rewrite? How hard would it be to built a client out of that? Because then you'd have instant deploy to Winmac, and support is growing for Android and MacThing compilation. Right now there's no UI framework for Android so it'd be Java bindings, but using the Golang backend for all the heavy lifting.

May I ask: is there any plan to extend the DHT with some sort of Pub/Sub, and with the NoStore BEP for mobile DHT nodes to indicate they shouldn't route or be contacted? Performance and speed are crucial for a microstatus network, but preserving bandwidth and battery are crucial for mobile users. Measures to speed up fetch operations and reduce DHT traversal are necessary, suggesting either hosted routers or pub/sub over DHT?

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

[Miguel Freitas]

On Mon, Nov 16, 2015 at 11:10 AM, Cathal (Phone) <cathal...@cathalgarvey.me> wrote:

Didn't someone make a Golang router rewrite?

Not really, it is just a network crawler + dns server.

 

How hard would it be to built a client out of that? Because then you'd have instant deploy to Winmac, and support is growing for Android and MacThing compilation. Right now there's no UI framework for Android so it'd be Java bindings, but using the Golang backend for all the heavy lifting.

I don't see much gain in rewriting the whole core code instead of focusing on the missing pieces (UI + installer + packaging etc). The twisterd is already ported to android and it works solid (some strategy is required to putting it to sleep so it doesn't burn the battery, but this is very minor).

windows port of twisterd also works solid (i've already provided virtual images preconfigured to perform cross-compilation). same for mac port. this is definitely not an issue, but if one wants to reimplement the whole thing... i'd never try to stop him ;-)

May I ask: is there any plan to extend the DHT with some sort of Pub/Sub,

I've never thought of that, so from my part, no plans.

 

and with the NoStore BEP for mobile DHT nodes to indicate they shouldn't route or be contacted?

Is there such BEP for bittorrent?

 

Performance and speed are crucial for a microstatus network, but preserving bandwidth and battery are crucial for mobile users. Measures to speed up fetch operations and reduce DHT traversal are necessary, suggesting either hosted routers or pub/sub over DHT?

Sure. These are all very important issues.

I think that not joining DHT for storage must be easy (from a cursory thinking, i don't know if any extension would be necessary). Then there is the issue of downloading the whole blockchain, we may surely do better for mobile.

regards,

Miguel

 

[1776]

Following up on this, the Android and iOS clients are critical.  Native gui, packaged up, open-sourced.  Easy for non-techies to install and use.
What do you all think about crowd funding development of a lightweight client for iOS and Android?  I would gladly contribute to that, I suspect others would to.  Raising thousands of dollars/euros wouldn't be difficult.  What about BountySource?  (vs GoFundMe or one of the others).

Thanks!

[1776]

One followup:

If we were to set up a bountysource for iOS and Android, it appears that it wants a github URL for specifications.  If we can set up the specifications for it on there, it will be easy to setup a bounty.

Would it make sense to make the issues under the main Twister github page?

[Amr]

Hi

I am an iOS developer and would like to help building the iOS client for Twister. Did anybody already started that effort, or should I start a new repo in github?

[Miguel Freitas]

Hi Amr!

The twister-core has been ported to iOS already, as you may see in this changelog here (merged):

https://github.com/miguelfreitas/twister-core/pull/339/files

There you may also find complete build instructions.

I'm not aware of any iOS client frontend, so please yes: feel free to start a new repo!

regards,

Miguel

--
You received this message because you are subscribed to the Google Groups "twister-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to twister-users+unsubscribe@googlegroups.com.
To post to this group, send email to twiste...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Amr Aboelela]

Ok

I will start a new one (God welling). I need help though developing a light-core for Twister which doesn’t run as a Bitcoin node but just as a wallet. And any needed user info by the client we should be able to get it, not from a centralized server, but from the DHT network. Now if the DHT network doesn’t have all the data we need then we should find a mechanism to duplicate the needed data from the blockchain to the DHT nodes.

Sent from my iPhone

To unsubscribe from this group and stop receiving emails from it, send an email to twister-user...@googlegroups.com.

[Amr]

here u are :)

https://github.com/amraboelela/Twister-iOS

To unsubscribe from this group and stop receiving emails from it, send an email to twister-user...@googlegroups.com.

[Frédéric JOUVIN]

Hello Miguel,

Where can we find documentation on how to use the experimental functionnality allowing to add attachments to posts with bittorrent magnets ?

Kind regards,

Frédéric

[Miguel Freitas]

On Tue, Jan 14, 2020 at 2:37 AM Frédéric JOUVIN <frederi...@gmail.com> wrote:

Hello Miguel,

Where can we find documentation on how to use the experimental functionnality allowing to add attachments to posts with bittorrent magnets ?

It is documented here:

http://twister.net.co/?p=589

regards,

Miguel