greengeek wrote:If such support came in the form of a 'piece of donated hardware' do you think this would benefit Puppy progress,
Yes, it definitely will.
or would such a piece of hardware be too out-ofdate as soon as it was available?
All things become obsolete sooner or later. But isn't it puppy's objective to breathe life a bit longer for them?
If there is benefit in this idea can you suggest some of the specs or features of the hardware that would enable developmental progress?
I can't, because the answer is "it depends". If you donate 86duino board then perhaps Puppy can be made to run on that board (with difficulty
). If you donate quad-core Core i7 with dual graphics and usb3 perhaps you can get puppy to run on it. The point is, the hardware can be used for many things:
- test/troubleshooting platforms
- build/compile platforms
- etc.
eg Making Puppy handle usb 3 - or has that already been solved?
usb3 is handled at kernel. Once the kernel supports usb3 then as long as puppy use that new kernel, usb3 is automatically handled.
A better example probably is bluetooth. Bluetooth support needs both *kernel* and *userspace* support. The kernel part is easy, the userspace one - not so.
Large RAM sizes - or is that issue sorted?
Large ram is solved by using PAE kernel or 64-bit kernel. Large RAM however helps to make build/compile process faster (a lot faster).
Recent wifi interfaces - or are they just too varied?
They are too varied. For challenge, try to get one with broadcom wifi - but don't cry it you can't get it to work.
Recent video cards - or are they also just too varied?
Too varied. The most important thing nowadays seems to be "dual-graphics", and this, like bluetooth, requires both kernel and userspace support, so it is challenging.
Of course, when speaking about video, it depends, one can try to install multiple video cards and trying to get all of them display at the same time; and/or making a giant screen composed of multiple outputs ... the point is - what is it do you want puppy to support?
Or would the value simply be in the fact that the hardware was powerful enough to crunch the numbers for faster compile times etc?
Yes, it is one of the things that helps.
The conclusion is: give the hardware according to the objective to be achieved. If you want puppy to run on certain hardware / new hardware, then give that. If you want puppy to run on obscure hardware, then give that. If you want to ease the burden of compiling, then give that. Sometimes you can achieve multiple goals within the same hardware.
Or, it can be given as $$ to free someone's time to do full-time development (like what Barry does), say for 3 months or for a year. Of course, before this happens, there must be objective measurable milestones to make sure it isn't abused. "Linux Foundation" originally existed to free Linus' time so that he can focus on Linux and nothing else. It works. (In fact it worked too well that now it can pay not only Linus but many others too). That being said, this may be too ambitious for this community