Hi, let me add my 3 cents here. On 15.02.2021 11:07, Greg Kroah-Hartman wrote: > On Mon, Feb 15, 2021 at 10:53:16AM +0100, Andrzej Pietrasiewicz wrote: >> W dniu 14.02.2021 o 03:17, Paul Wise pisze: >>> On Sun, 2021-02-07 at 07:28 +0800, Paul Wise wrote: >>> >>>> All the standard benefits of the GitHub organisations feature and >>>> collaborative maintenance in general >>> >>> Since the linux-usb GitHub organisation isn't yet available for >>> collaborative maintenance of Linux userspace USB gadget projects on >>> GitHub, we will create a linux-usb-gadgets GitHub organisation and can >>> move gadget projects to the linux-usb org when it becomes available. >>> >> >> Any prospects of it "becoming available"? > > I still fail to see what this is going to help with here. Are usb > userspace projects going to somehow get loads of more developers > somehow this way? What is preventing that from happening today that > dumping them all in a single project going to change? Partially it's my fault. I'd love to help with further libusbgx development but in my work time I'm dealing now with some other challenges and additional duties in private life does not allow me to continue active development. I tried to at least review and merge pull reqs but the context switch of my brain turned out to be to heavy. > > Am I now somehow the arbitrator of what is, and is not, a valid project > to join? We already have competing libraries scattered around, lumping > them all in a single location isn't going to change that problem from > what I can tell. I already had this discussion with Matt Porter. We agreed that libusbg is obsolete and libusbgx is the direct ancestor for it. If you take a look into commit history you will see that libusbg is just a small piece of libusbgx history. Now when it comes to the other projects. My company actively uses two of them in tizen: - libusbgx - gt gadgetd has been developed by us but at the end of the day it was dropped and the functionality has been simplified and merged to the other system daemon that we have - deviced. For the two that are used in tizen. They are just there and serve their purpose. They are definitely not feature complete but the use case in which they are used there pretty much is. Based on what I see in github traffic I expect that the three projects that definitely have some traction and interest are: - libusbgx - gt - mtp-responder > > So, what problem would this solve that the added maintenance burden by me > and others would be worth it? > From organizational perspective I believe that keeping those 3 together makes sense. Personally I don't have any strong preference under which github org they are going to be published. For me the most important is to have them easily accessible to people and have someone onboard who can help with the maintenance. Best regards, -- Krzysztof Opasiak Samsung R&D Institute Poland Samsung Electronics
