Open to collaborators?

[0xdevalias]

Hey @jehna, was wondering if you were open to potentially having me join the humanify project as a more 'official' collaborator? Totally understandable if you'd prefer not, but I was thinking it would allow me to help triage/issues better/etc.

I also have a few ideas/local WIP refactors to clean up some of the CI stuff, and then maybe also refactor some parts of the codebase + add new features/etc (which I'll probably do as PR's regardless of the outcome of this)

No rush/pressure, but let me know what you think when you get a chance :)

[0xdevalias]

Some notes/context from a recent email conversation, RE: my interest in potentially forking humanify to be able to more easily work on things (but not quite having the capacity in my schedule to focus on it just yet):

Yeah, I similarly sent them a LI request a little while back, opened an issue about collaborating, and even more recently tried a direct email; and no response from any of them. They did end up merging a couple of my tiny PRs back in early February (1, 2), but not this slightly more involved/untested one in late February; and haven't seen anything from them since. From a quick skim across their blog/socials/etc it seems they've been pretty absent from there for quite a while now too. They've still been somewhat active on GitHub though according to their contributions, mostly in private repos, but a couple of public ones too (1, 2). This was the email I sent a couple of days ago:

Heya,

Forgive the direct contact; this will be the only email I send unless I hear back from you positively otherwise.

I noticed that you haven't been very active on the humanify repo for a while now; and from a quick skim of socials/etc maybe not there either; and so I wanted to reach out and see if you were open to adding collaborators to the humanify project?

• Open to collaborators? #358

I have a number of ideas of refinements/improvements that I'd love to dive deeper into (as soon as I clear my current little backlog of unrelated tasks); but I'll admit that the lack of activity on the repo has made me a little hesitant, as I don't want to head in a direction you wouldn't be comfortable with, or have the PRs just sit there unmerged/similar.

Anyways, hope things are well with you; and sorry again for the direct reachout.

Cheers

Glenn 'devalias' Grant

RE: forking; yeah, that's basically been in my mind for a while now, though I have been pretty busy with other life things (travel, catching up on all the things I ignored while travelling, etc) so hadn't got around to it yet. In reality I will likely be somewhat busy for the next few weeks or so while I get back on top of everything, so I wouldn't be able to put too much focus into this till that is done; but I am definitely interested in exploring the idea of forking the project and working to get it cleaned up a bit more + get some new features/etc landed. (There's also another semi-related side project I have in mind that I hope to be able to put some more solid time into once I get on top of everything else too; but that is far more at an early R&D sort of level of things for now)

And then this followup:

this love is why you should be co-maintainer or just do a fork

Yeah, I'd be happy to just help out on the current repo if @jehna was open to that; but given the lack of contact/etc, fork is probably the next best thing.

[skilbjo]

+1 allowing @0xdevalias to co-maintain this project w @jehna , or @0xdevalias if you want to start the fork, i will co-maintain with you.

assuming neither, i will just do a hard-fork myself, this project is too cool and exciting to leave unmaintained.

[0xdevalias]

if you want to start the fork, i will co-maintain with you

@skilbjo I'm open to doing this, though might not have the capacity to do so immediately.

As I raised in my original email, did you have thoughts about the structure of the fork (eg. as an organisation, etc?):

Interested to hear your thoughts on future directions for the project. Are you interested in collaborating/working on the Humanify fork with me? Or more just hoping to see the project get some more direct love/updates? Did you have any particular thoughts on how you would envision the fork? Like in a new GitHub org for it, or?

---

Also, if we are going to fork, I'd also be interested to hear if @j4k0xb (https://github.com/j4k0xb/webcrack) / @pionxzh (https://github.com/pionxzh/wakaru) have any interest in being involved in it in some capacity (either directly at a co-maintainer sort of level; or more casually/indirectly); as I suspect there is some crossover potential with webcrack in particular (that humanify builds upon); and it's upcoming plugins feature/etc.

I suspect there is probably some re-architecting that could be done to make humanify nicer UX / DX overall (like it's always mildly bugged me the way the CLI splits local / remote LLMs / model downloads across separate commands/etc; as one example)

[skilbjo]

yea, if i hardfork it i would redo the project organisation in my favourite style, similar to what i've done in https://github.com/xhrdev/examples

ie

src/ folder
test/ folder
use jest
use the eslint settings i am familiar with
use a Makefile

i don't have any opinions on nicer UX/DX. i think you would be perfectly suited to make those improvements. i just see such good PRs being left unmerged, and it bothers me.

[0xdevalias]

@skilbjo Hrmm.. fair enough. Makefile is somewhat non-canonical for a node project IMO.

With forking it, I was thinking of making a new org and having the repo in that; potentially as something that could act as collection of web reversing tools like this.

---

i just see such good PRs being left unmerged, and it bothers me.

@skilbjo Understandable.

[brianjenkins94]

Just wondering if anything came of this? I would really like to have better handling of name conflicts.

#330

[0xdevalias]

Just wondering if anything came of this?

@brianjenkins94 The short answer is that despite trying to contact them in multiple ways over time, and despite them having recent GitHub activity, there has been no response from @jehna about this at all.

Based on that, @skilbjo and I have been talking about forking the project. Their plan was to hardfork it with a new name; I probably would have just forked it as something like humanify-ng and kept it closer to the ability to merge the projects back together in future if @jehna ever comes back to us on this.

But so far, to my knowledge / last time I checked in, I don't believe @skilbjo had been working on the fork so much due to other commitments; and similarly, I unfortunately haven't yet had capacity to look deeper into it myself.

[skilbjo]

yup @ that's the current status !

both @0xdevalias and i are hoping @jehna returns and just maintains this project, but in case he doesn't, i am keen to build a very similar project (and mostly inspired by) humanify, that i am calling Cipher. i think @0xdevalias would be a great co-maintainer on this with me, but i want to get it a little more built out before i invite him to the project.

i was able to get mostly what i was looking for out with aistudio.google.com in the meantime tho, hence why it hasn't been much of a priority for me. however i still am keen to do it!

[datoslabs]

Hi,

I recently came across this project and started experimenting with some examples. I've made some enhancements with my local copy to address some issues, more specifically:

1. Ability to resume after failure/interrupt, at outputDir file level
2. Added timeout and retry local llm prompting after the process gets stuck on session.promptWithMeta call; when this happens, the process is still running and consuming CPU/GPU resources but the session.promptWithMeta call awaits indefinitely. It appears that the LlmContext needs to be disposed and reinitialized after a few hundred inferences, this allows the process to continue when processing large files.

I think there's a lot of potential for this project. I would be happy to open a PR or create a simple fork as @0xdevalias suggested to house these changes while maintaining the ability to merge/rebase with this project in the future when @jehna returns.

[0xdevalias]

I would be happy to open a PR or create a simple fork as @0xdevalias suggested to house these changes while maintaining the ability to merge/rebase with this project in the future

@datoslabs While it's unlikely to get merged any time soon; for visibility it would be cool to make a PR with your changes. Then we can pull them in more easily when we fork/clean things up.

[datoslabs]

@0xdevalias I will spend some time to clean up my code after more testing before creating a PR. My main concern with continuing to create PRs here without the ability to merge into main/dev branches is that over time the project becomes unmaintainable and drive away contributors. A simple fork with active maintainers can ensure PRs get merged in a timely manner before new code/fixes are added.

[0xdevalias]

A simple fork with active maintainers can ensure PRs get merged in a timely manner before new code/fixes are added.

@datoslabs Sure, I agree; but are you intending on maintaining that fork? If not, and it just ended up being a fork with your changes on it as well; I feel like that would just fragment things more than an unmerged PR here would.

[datoslabs]

I will certainly contribute but I don't have the capacity to maintain a fork alone. The main reason I commented here, instead of creating a new post with the changes I have made, was to gauge if more are interested in a joint effort to maintain a fork that allows others to contribute to this project and move it forward.

[skilbjo]

i started a hard fork but it's immature, currently private. i'm travelling now and won't really have stable housing + quiet time until later this year (i'm in europe for the summer, seeing various founder friends), but i'm happy to invite @0xdevalias and @datoslabs to it and we can build it out in private until we think it's ready for public release. i think in particular @0xdevalias would make a great co-maintainer with me.

i will be able to do a bit of work on it while traveling, but can't become the main steward until things settle down for me.
once it's established, i would be happy to be the main maintainer of the project for the foreseeable future (3+ years), as it's highly relevant to the work i'm doing @ xhrdev (a commercial side project of mine)