Is this repo dead?
luntik2012 opened this issue · 11 comments
Looks like it.
Did you guys try to build it? is it working? I am on debian testing.
It does still seem to work @awalis. I'm building master on my archlinux just fine. It didn't work on my personal mega account but I've been able to fix the issue by editing the source. Won't be submitting it tho since no activity and all recent pull requests seems to be ignored.
@tYYGH oh yeah, I did fork it after all, committed the fix too
Just a thought… Misc. people already forked because of math.h/cmath. Apart from that, we already have:
- your fork (@Unknowny) for an API fix,
- the fork from @vividcatastrophe, apparently for a Linux systemd fix,
- the fork from @henning-schild, for bug fixes in the code,
- the fork from @gpiemont, which seems to target NetBSD compatibility.
That’s a lot of nice work, and it’s a loss for everyone, that each goes their own route… Why wouldn’t you guys coordinate? For example, @gpiemont’s fork being apparently the most divergent from @matteoserva’s reference, maybe @gpiemont’s fork could become the new reference (if he/she agrees), and they you’d only have to send pull-requests to him/her?
More over, there are older but still interesting fixes by @malexan and maybe others, that should not be too hard to merge…
Cheers,
Anyone want to fork and maintain to be a base for accepting PR ? in meantime until @matteoserva back.
Can you @tYYGH ?? your sumarry above is awesome, i just cherry pick them to build this
sorry if i just suddenly butt in.
Hi all,
yes my fork targets netbsd compatibility.. but unless you're interested in netbsd software porting or similar, don't take it as a reference. It's undocumented and it's still based on an ancient MEGA SDK and it definitely needs to be ported to a newer one (or the latest).
Nevertheless it works fine under netbsd (tested on netbsd-7 branch).
But as I said, it's the main core that needs to be rewritten: memory management, download manager, etc..
/me agrees on waking up @matteoserva
Cheers
Hey guys, just a note, I recently discovered https://github.com/ncw/rclone (which made me a bit depressed for making a good bunch of basically identical ad-hoc solutions). They got mega integration and virtual filesystem mounting capabilities also. If this project lives on it might be beneficial to share code / knoweledge and otherwise.
I think that mega integration is a bit difficult nowadays.. It rises a lof of problems when used from non-compliant software. The only one that works smoothly is megaCMD, along with megasync and the browser extensions.
The login phase seems to be the most problematic one: it takes too much time in @matteoserva's megafuse and it fails for the wrong reasons on this rclone.
Think about it before loosing time on this. Most of the issues depends on the new storage policies adopted by MEGA and only in part from the SDK implementation.