:IF: I AM IMMUTABLE! Blogging with IMIM

you’re killing it :kitchen_knife:

4 Likes

Great work Traction :clap:

I see you’re using Pointers as well and wonder if you’ve seen them updating ok on mainnet?

I’m only seeing the counters stuck at zero but they work fine for me locally so I’m still assuming it’s a mainnet issue.

4 Likes

I’ll point out again that for me pointers do update on main through python

But possibly it’s just working so well in python because the language is awesome :face_with_monocle:

6 Likes

Yes, they seem to be updating fine. I’ve not stress tested it, but each time I’ve tried, it’s worked.

Iirc, 1 register update got a bit confused and it took another to remedy it (as the client library log suggested it would), but it seemed pretty smooth.

maybe it’s not only that python is so awesome

AntTP does use the pointer_update function while the history datatype in dweb uses explicit state management and pointer_put
so slightly different mechanics

4 Likes

Good spot!

pointer_put Only use this when you know what you are doing. Else use pointer_create and pointer_update. Manually uploads a pointer to the network for a payment. Returns the total cost and the pointer’s address.

I read the first bit above and figured I’d use pointer_update instead! :sweat_smile:

The pointer_update is free too, which is an interesting difference vs registers (and maybe pointer_put? :thinking: ).

1 Like

Good detective work, thanks. My code is modelled on Registers so I expect they have the same issue.

1 Like

funnily I dug a bit further and remembered that you mentioned this … registers interestingly (by now?) are using only pointer_update and don’t do puts anymore :open_mouth: … after some circles I’m now pretty sure there is a bug in state management (that is pretty irrelevant for the dweb publish use case) but still don’t understand why it would fail xD … I wrote you a wall of text on discord xD

2 Likes

They charge for the version that doesn’t work :thinking: :rofl:

Thanks fellas, I was closing in on this but @riddim to the rescue :folded_hands:

I’ll fix this today and everything should be working to at lest reasonable level. :folded_hands:

4 Likes

Time for a little teaser!

Yes, that’s right - adding and updating new blog articles is almost with us!

There is a bit more plumbing to be done, but I successfully created a new blog with an article through the app today, without a CLI in sight.

Hopefully, I’ll get something to play with uploaded this weekend.

16 Likes

I’m not prepared to leave terminal land yet :shaking_face:

wohoooow!

5 Likes

After fighting and winning a parameter war with ant file upload (--max-fee-per-gas 40000000 is special, hot, sauce!), I’m pleased to usher in web based article publishing an IMIM!

You will need the latest version of AntTP, but other than that, you should be good to go - the pointer to IMIM now points at the latest code base, so you will automatically see the new app version.

Moreover, there is a default ‘imim’ bookmark set in AntTP, which whisks you straight to IMIM, without any long XOR/pointer URLs.

So, if you have your Autonomi / AntTP browser ready (i.e. browser with AntTP as a proxy), you can go straight to this URL and see my blog and the new publishing tab: http://imim/blog/traktion-blog

So, the perfect opportunity to munch on some dog food too! I’ll let IMIM tell you more about it…

http://imim/blog/traktion-blog/article/say-goodbye-to-the-cli.md#article

It still has a few rough edges, but it’s starting to come together nicely!

EDIT: Just to note, it looks like some of the CSS wrapping is a bit abrupt on the new article. I think it is the bullet lists, but I’ll tweak it into shape tomorrow ! :sweat_smile:

8 Likes

I typically use 10 million which is currently the lowest possible. Unless the GAS fee is high this works and ensures no load when fees are high

3 Likes