No; CPUs didn't /have/ to do that. MIPS toyed with both models for a while - initially MIPS was like "we don't interlock pipeline stages, so programmers need to be smart." Then the R4000 came out that attempted to implement that, and it was .. complicated. So it got reverted.

Not all CPUs are like Intel CPUs (which aren't all like earlier intel cpus, which aren't all like 8080s, etc..)

The reason is because the manufacturer gives you some old version of linux that they have included as their "base" reference design OS, and .. people just build on that.

It's purely inertia.

No, they used linux because that's what Atheros gave them as a base for the Atheros reference AP implementation.

Please don't make stuff up.

...

Did you just try web forums? Did you try engaging people on one of the FreeBSD mailing lists? It's odd that repeating a crash like you did wouldn't lead to a quick solution.

I mean, unless it's a realtek NIC or something, we're pretty good at fixing repeatable crashes.

(adrian@freebsd)

.. because you've just exactly outlined why it's a bad idea.

If you as a service provider can afford to pay, you will. If you're a new service trying to get a foothold, now you have to meet those costs too if you want to play. Those costs aren't fixed and behind general transit/interconnect/infrastructure fees, they're behind whatever-the-telco-decides-is-a-good-match.

Australia went through this. We have and had phone plans that have/had free access to certain services and not others. Guess what? It sucked. It may be great if you upload lots of photos to facebook, but it means you can't at all start a new service that competes without having to get individual agreements with individual telcos and service providers. You'd have to negotiate those deals individually and your business will only exist as long as they don't alter the deal. They can then alter the deal just to you but not to their favourites.

It doesn't work the way you outline it.

I was working in the first dot com boom during 1998-2001. Now, I was working in Amsterdam rather than the US, but I did get to feel exactly how screwed up this situation got. And looking back at it, this article does re-iterate a lot of those points quite clearly.

The people that succeeded were for the most part the ones that put in long hours, were ruthless about achieving their goals and cared not about things like "work/life balance", "emotional stability", "health concerns", etc. Whenever I came out to the US to talk shop with other internet infrastructure people, they were working long hours, ignoring what the industry said they could / couldn't do. There were women in tech, but they weren't the programmers - the ones I met tended to be algorithms people, data scientists, etc. They weren't in the meat grinder of bashing out C/perl code.

The article covered the long hours, it covered what happened when things went pear shaped, and it did a pretty good hatchet job on the kind of focus and ruthlessness you needed to get where you wanted to go. It was amazing to watch and now a little scary. Then the dot-com bust happened and people lost everything. Plenty of people I knew said "fuck it" and left the industry. Those that stayed either made their money, or they were just suckers for loving their jobs. They didn't have strong personal relationships with others. They just loved kicking ass and taking names in their work career. That sometimes worked out for them and sometimes didn't.

A lot of the people I knew in the tech field did just leave and look for something more stable. The people that stuck it out were homeless, couch-surfing, living with family/friends, existing wherever they could just to get over the sheer loss of everything. Not everyone is cut out for that level of destitution and dedication - eventually they'll snap and go off to something more stable.

This field is terrible. It chews you over and spits you out. If you're lucky then you make a bunch of money and save a bunch of money. Plenty of people working in tech and living in San Francisco aren't even doing that. We don't necessarily churn out people who are risk takers out of university - heck, churning out creative thinkers just became an "in vogue" thing again with this whole maker faire mentality that's happening nowdays. But when the thing crashes again, you'll see the same cycle - those who are willing to risk it all and live hand-to-mouth from wherever they can will do it. Others will go find whatever is safe and stable and start life again from there.

Now, is that gender biased? Maybe. Someone has to go do a little more research to figure that out. But from what I saw, there were a handful of women that stuck through that and came out ahead. Most that I knew just gave it in and went back to school, moved in with parents, or decided to stop work and have babies. The guys seemed more happy to take the risk again and again and live hand-to-mouth.

There's lots to fix. We have to stop being insensitive asshats. part of that is institutional - the brogrammer culture is strong here. Part of that comes out of all of the stupid stress and anxiety that litters this community. It's hard to pay attention to how you live, how you interact, how you make others feel, how to communicate well and well, how not to be an asshole if you're always stressed out, anxious and sleep deprived. add in a bit of being shouted at and some threats about your job security and .. well, you just stop giving any fucks. Part of it is no constant exposure to dealing with other people and a focus on your ability to churn out code - your job doesn't tend to want you spending time each day to improve yourself in all ways - it needs to be work relevant, and hey you have that deadline that just appeared? Eww. It's good to see people standing up and calling out bad behvaiour. it's good to see that some communities are sprouting up and eschewing shitty behaviour. But I'd really like to see the stress, anxiety and hours drop as well as a focus on people interaction. My 20 year old self gave no interest to any of these things. My 35 year old self .. suddenly realises that it's pretty fucking important.

(I really should write the damned 7260 driver. Ugh.)

I didn't realise our cp didn't support sparse files. Would you mind filing a FreeBSD bug about it? We use Bugzilla now, it's not AS bad as GNATS.

Thanks!

-a

Yeah, I can explain that.

The current setup replaces the repo each time - and if a package fails to build, all the dependent packages can go away.

I've had chrome disappear from my freebsd-head install (which i manage with pkg from the test repositories, so as to dogfood our own stuff) and it's generally been because it's been marked as having a security problem and the port wasn't built.

Hi, adrian@freebsd.org here. If you've had issues with suspend/resume and poor battery life then please drop us/me a line. That's something a group of us are slowly fixing.

-adrian

No, the problem is that we don't have as many active wifi developers as other projects, but the desire/need is still there.

I now have a couple of people helping me chase down intel driver bugs and implement / port the 7260 driver from Linux/OpenBSD.

-adrian, typing this on a freebsd laptop with a centrino wifi adaptor working just fine.

Hi,

I'm the wifi guy. Your post is misinformed. FreeBSD developers use their OS on non-server hardware. Everything you said above works. Even the wifi.

https://twitter.com/encthenet/...

You're very welcome!

-adrian

Not entirely - it's an easier problem to solve when you're parking a car that takes up a single spot. Now, imagine you have a trailer count that is between 0 and 1024 parking spots wide, and breaking your trailer up into pieces (and then reassembling it!) is feasible but it takes time to do that.

That's why it's not that simple.

There was and likely is some hardware that does it.

It's also easily DoSed.

We found this out in the 90s and early 2000's where people would .. well, try doing internet routing with Sup-1's.

Except people can, have and will deaggregate IPv6 space to do Traffic Engineering.