Godammed SDN

[Veef]

Oh man baks kun got told by shepper on conventional war

[Flagg]

I'm out of the echo chamber. What will I do with my time?

Smoke more crack?

[Darksi4190]

Oh man baks kun got told by shepper on conventional war

I like how 100,000 Humvee equivalents are somehow a gigantic cost sink and job creator on the same level as the massive cold war era production of tanks, aircraft, ships, missiles, ETC.

[Veef]

So are Crown Vics

[Oxymoron]

Oh man baks kun got told by shepper on conventional war

I like how 100,000 Humvee equivalents are somehow a gigantic cost sink and job creator on the same level as the massive cold war era production of tanks, aircraft, ships, missiles, ETC.

I like in particular how, by trying to explain why guerilla wars which aren't fought like conventional war (which was one of Bakk's point) are still a giant money sink ; how he only ended up inadvertently reinforcing Bakk's commentary about the American Militaro-Industrial Complex.

[Shroom Man 777]

gods being away from TEO feels so liberating

i can remember the first days when it was quite tough and i had to tell myself no

but now, oh yeah

[Flagg]

That South Park thread is a lol. Apparently thinking obscure shit isn't funny when satirized makes me a bad person.

[xon]

I forgot how much sd.net really don't like having technical questions answered.

linky. USB boot has a rather long list of drawbacks, and since Windows needs to work with more than 5 different types of hardware some cave dwelling linux kernel developer supports, not rolling out a feature cos the majority of the user base can't use it sounds like a good idea.

[Zod]

is there any real value to booting from usb for the average user anyway?

i'm just not seeing it for anyone who's not fat

[Oxymoron]

AFAIK, USB-Boot has the same usage as booting from a live-CD [1], with the bonus of the "system" being Read/Write instead or Read-only.


[1] : basically, booting into a computer with difficulties to boot on its own. For example, that's useful in case of virus-contamination, but in this case you'd prefer a read-only live-cd, so... *shrug*

[Zod]

so basically no.

[xon]

is there any real value to booting from usb for the average user anyway?

Hell no, which the boggling at all this raging over a feature they, almost, will never use. I'm not really sure what it is about my post which inspired such rage from a few posters in there.

i'm just not seeing it for anyone who's not fat

It only has a very few corner cases in which it is useful.

[Zod]

Hell no, which the boggling at all this raging over a feature they, almost, will never use. I'm not really sure what it is about my post which inspired such rage from a few posters in there.

atheos is a fuckwit, so there's at least one explanation

[xon]

atheos is a fuckwit, so there's at least one explanation

Yup.

[phongn]

I forgot how much sd.net really don't like having technical questions answered.

Most people on SDN who think they know something about computers don't actually know that much?

[Zod]

I forgot how much sd.net really don't like having technical questions answered.

Most people on SDN who think they know something about computers don't actually know that much?

i think they just have a massive sense of entitlement and no social skills

[adr-admin]

remember that one time i got in a flamewar with nitram because i was obviously a retard for saying win95 wasn't based on dos?


i stopped even looking at g&c a long time ago

[adr-admin]

also "linux does it"

ummmm no it really doesnt

they work around a returning device in their own way..... which breaks fucking horribly in the situations windows is able to deal with

i remember one time we were trying to get a usb network thing to work on linux and the drivers were painful but got past it

and then after doing some switching (this was a while ago and i don't recall the exact steps) it stopped working

not cuz of hardware

not cuz of drivers

but cuz some config file mixed up the recognition


took hours to figure it out

[adr-admin]

or oh god linux booting off certain devices

it seems like it is working

then kernel panic cuz of some scsi shit

this was in VM WARE and the emulated hard drive wasn't reading as a standard hard drive

wtf

"oh you have to recompile the kernel with scsi emulation irq reassignment blah blah" i dont recall it was technobabble

then you have the stupid crapwhere the same ide device gets a new name

and then it kernel panics because it can't mount the root filesystem


son of a bitch

im glad i don't deal witht hat much anymore

[Oxymoron]

Why does it seems that in real-world engineering, conceiving things in very conservative ways, taking minimums risks and making sure things will works in the most adverse condition is something every engineer should strive for

while in software engineering it's all shortcuts and custom "cook-book" solutions ?



Are all programmers lazy slackers, or is it that software engineering is such a young field (only 30-40 years) that the idea of a standardized/harmonized work methodology at the Profession-level (and not just at Corporate-level) still has difficulties imposing itself ?

Is it all the fault of the "Open-Source" crowd, made of happy amateurs, or is it something else ?

Or am I just spewing bullshit ?

[Zod]

software doesn't really have any hard and fast rules like physics or chemistry

[Oxymoron]

That and the fuckton of differing programming languages, hardwares, etc... ; and the really quick evolution of the field. It sure doesn't help.

I was more thinking in term of general work methodology like UML or things like that.

[xon]

remember that one time i got in a flamewar with nitram because i was obviously a retard for saying win95 wasn't based on dos?

Yeah, all MS-DOS was a highly interactive boot-loader. Win95 technically didn't need it to exist but it wasn't worth replacing.

Why does it seems that in real-world engineering, conceiving things in very conservative ways, taking minimums risks and making sure things will works in the most adverse condition is something every engineer should strive for

while in software engineering it's all shortcuts and custom "cook-book" solutions ?

Software development can crudely thought of searching through the list of posible permutations of programs as bit-strings and selecting the approximately working program for a goal. Any sane person would reject this as a valid design strategy, but what the hell do you do when you don't actually know any sane solutions or even methology to develop them?

Engineering has well understood physics and limitations which shape how you do stuff, but software just doesn't have that. You are left with taking known working solutions which kinda answer the question you want, and it's more cost-effective timewise to hammer it into shape than rewrite everything from the ground up. But even those 'cook-book' solutions require non-trivial investment of human thought to get working.

Or am I just spewing bullshit ?

Nope, and that's the sad part.

software doesn't really have any hard and fast rules like physics or chemistry

And this leads to some amazingly horrible ideas. Linux as a Microcosm for what is wrong with the cult of choice for example.

[xon]

I was more thinking in term of general work methodology like UML or things like that.

UML is largely worthless. It's an amazingly verbose 'langauge' which requires extensive(and expensive) tooling to work even vaguely right and has massive overheads.

The idea of automatic code-generation from design tools has largely been recognised as only useful during the initial design, and doesn't help for maintaince which is about 90% of most non-game software's budget (no really, software maintaince easily dwarfs development costs most of the time).

[Oxymoron]

My mother, which has she said has worked on "IT Urbanism" [1] (she has been retired for 5 years, now), told me that lately the work methodology was to begin from test units, and going up from there : you know your inputs and the outputs you want, you build you block, you test it, you debug it, rinse and repeat until it's proven to work, and then you move to your next block. Extreme Programming it's called, I think. She said it had good results, in term of quality and speed-of-development.
She was working at France Telecom's R&D department, though, so it may not be a work methodology adapted to every corporations...



[1] : from what I understand of it, it's the same distinction between an architect (building a single house) and an urbanist (arranging a whole city with numerous building which have to work together).