selsta: I just rebased my #8076 after the merge of jberman's #8556 caused some conflicts. Now the time for merging my PR has hopefully finally arrived.

I wonder how to best check such a merge process on GitHub, to make sure I really only changed things to correct conflicts and did not take the chance to sneak in something malicious

Because of the merge there are of course tons of differences now to the last commit of my PR, with over a dozen commits added to master

Is there a way to check only my merge-conflict related changes?

I am talking about this difference: github.com/monero-project/monero/co…69184b81a6a22100715ba1c81439043ce54

Just wanted to say thanks to all involved for pushing out another release! Your work's much appreciated. Now the testing starts :)

my hashes paste.debian.net/1254060

what, is it tagged already?

I'll build binaries in the evening

my hashes: paste.debian.net/hidden/f2ae0069

looks good

sech1: please upload them again to the p2pool website

rbrunner: will take a look once the release is out

luigi1111: please also gui merges + tag once you wake up :D

Ok, thanks

selsta I'll upload when I get home in the evening

ty

My hashes: paste.debian.net/1254086

My hashes again. First paste was going to expire in 24 hrs. paste.debian.net/1250489

s/1250489/1254089/

selsta p2pool.io/bin

sech1: thank you, have to wait for gui tag

Hi

hi

Jesus fuck, whoever made 672 API layers to reach wallet2 should be... told not to do it again -_-

Every layer with slightly differnt types, kill me ow...

And of course Qt just HAD to have its own string class, because using std::string is for people who don't flaunt their DIFFERENCE.

I'm gonna make my own just to spite the next poor bastard that reads this code...

Reminds me in TF, which has String from the game engine, TBStr from the UI lib... I think there's a third one hiding somewhere too...

Ah, tstring in litehtml. Though that one resolves to std::string (and, obviously, somthing else for windows, because fucking windows)

No way. One layer has *both* Qstring and std::string in different functions. Asshole...

What are you looking at?

Garbage.

The glue in feather wallet. Which I suspect was lifted from the monero GUI, but I haven't looked.

Yeah, but where is that garbage? I could end up designing the API for the Seraphis core wallet, and I am looking for inspiration ... or examples how not to do it :)

This particular one... src/libwalletqt/WalletManager.cpp

Yeah, that "Wallet API". It seems to be quite successful. The GUI, Feather, Monerujo all use it, as far as I know. Don't really see the attraction over "pure" wallet2.h, but maybe I did not look close enough.

The attraction seems to be "I can now juggle with one more layer". Number goes up ?

Look ma, without the brain!

OK, I'm being unfair here. Just annoyed at all the glue crap just to add a bloody function.

As long as I don't have to change something there, it's possible to ignore it. Do you have to change?

Well, depends what you mean by "have to" :)

Ah, yes, "add a bloody function" :)

Take it as a test of intelligence, and you passed with flying colors. Congrats. Other people would faint at the sight of the third string class :)

I took it as a test of my piss taking detector :P

I mean, I think I did my own string class like twenty years when I found C++. It was fun to do these little things. I don't think I ever made a lib with it though. Sorry if I did...

OK, victory achieved over that gunk code. Feather now has hidden TF messages :D

Pointless, but cool.

Well, display mode. I now get to add send code. Just for hte hell of it.

(and, for hte record, I have a cc_message_t class. I had to add *two* different versions, one that uses std::string instead of crypto::hash, because the first wallet2 api layer uses strings instead of crypto objects, and another with QString, because Qt is scared of standard strings).

This is ridiculous.

With all the conversions of course, when you go through the layers.

And the slightly diffrent prototypes.

Someone should make a javascript API for this. That'd probably reqiure some new js string type.

WASM. WASM is it today, I heard.

I still have problems to believe that you can compile wallet2 and crypto code to WASM, an that bloody thing runs then.

What, Another String More ?

Yeah. That's above my pay grade. I tried to look at it and failed miserably.

Wasted way too much time before I cut my losses :(

Do you build something like Feather Wallet, TF edition?

Yes.

Because why not :D

So it manages, what it is, simply "gold"?

Yes. It's really monero under the hood.

Why not - good question :)

I remember, yes.

Still waiting for the ICO.

reminds me a little of this PR monero-project/monero-gui #3965 (with ints this time)

was broken for 2~ years

Nice. Objects are comparable but different types so never equal ?

yes, it gave me a few extra wrinkles

always failed the checks and selected the default value

Hmm, I guess I should also complain about functional_tests being in python (that one might be my fault), so you're told about simple dead obvious typoes after one hour of runtime.

OTOH it means there are lots of tests. Glass half full.

soon