:: Re: [DNG] What do we want for ascii…
Inizio della pagina
Delete this message
Reply to this message
Autore: Simon Hobson
Data:  
To: dng
Oggetto: Re: [DNG] What do we want for ascii ?
Go Linux <golinux@???> wrote:

> This is putting the cart before the horse IMO. It would be nice to get the beta out the door before focusing on ascii. Any chance some of that energy could be directed towards the beta release?


I'm kind-of on the fence here.

Part of me is saying yes, look forward to what the next version should look like.

Another part is saying that there are priorities with the current release that would be a better spend of the very limited resources available.

Which way you go depends to a large extent on "what are the aims of the project ?"



IFF the aim is to take over the mantle of "GNU/Linux for the masses" - but without systemd then IMO some effort is needed into the infrastructure. I suspect I'm not a particularly unusual type of user - I'm considered a guru by the others, but in reality I "know what I've learned". I can configure Postfix, Apache, MySQL and a bunch of other stuff to build servers to do specific tasks. I'm familiar enough with Xen to now be running around 20 virtual machines instead of a stack of boxes. I understand the networking stack well enough to build routers with multiple interfaces, traffic shaping, and do pretty graphs. I'm fairly fluent in Bash
But compiling stuff has never gone beyond following a crib sheet of "make clean, make configure, make" and if it goes wrong then I'm stuck.

So pretty well all my boxes run standard packaged stuff, and (from memory) I don't think there's anything hand installed. That's actually a positive decision on the basis of leaving as few land mines for anyone who might have to pick up where I've left off to step on. Never rule out the proverbial accident with a bus.

Did I mention, I'm the sole "Linux guy", and refusenik Mac user, in what is essentially a Windows shop ? No, well I have now. When I started here, the DNS was on Linux (though I don't know why, I think there was another "a bit Linuxy" guy before I started - I had to fix it on the morning of my first day here !), and a web server for those things that needed it. What I have now is by having been able to build things that work - or work better than something built with Windows*.

Anyway, part of selling "Linux" as an option is using something that's "acceptable". Debian is fairly easily seen as something that's been here for some time and has enough momentum behind it that it looks like it'll be here for a while. That's also how RedHat got where they are - by selling something that people can seel to mangelemnt - as in "see there's this company we can shout at if it doesn't work" even if that is as ineffective as it is shouting at Microsoft !
As to Devuan, this is not meant to be disparaging or insulting in any way, but the current state of the project comes across as very much a small group of people with very limited resources, and packages served from "various random servers". I can tell you are all keen and motivated people, but manglement likes to see "proper stuff" - and at the moment I would not try and sell Devuan to them.

It's a hard task, and Debian has had a couple of decades to build their's, but IMO a key target has to be to have the infrastructure in place - or at least the appearance of it - so that manglement can see an organised project, with repositories, a bug tracking system, etc like all the "proper" distros have. "Image" is important.



* When I started, we had an old mail server (iMail on an NT4 server). I had nothing to do with it as I've made a conscious decision to avoid Windows as much as possible. It got flaky, was inflexible, and it kept getting hacked. As a "temporary measure" I knocked up a box with Debian Sarge, Postfix, and a few other bits including a package called Postfix Admin - which outperformed the old box and offered more features.
It ran 8 years and I've only just retired it having built a better setup with similar packages but all newer versions !