I've been on a couple of projects where there was a lone woman (either developer or QA), and they were always treated as just another team member. OTOH, both times we were all on new teams, so there was never that awkward "new kid on the block" phase.
five additional people earned commit priviledges, but where not made to agree to these "new" criteria
If you have someone who is an alcoholic, you would require them to agree to stay off the sauce, but you probably wouldn't require it from anyone else. Technical arguments on developer mailing lists are common, but if they get so heated that developers leave or switch projects, then something has to be done.
The only "workable" solution that was offered was that Theo could pass his diffs on to the port developer and let him merge them. Basically it was a set of conditions that nobody would agree to.
You haven't contributed to many OSS projects, have you? That's actually the way most contributions from new developers get made: you pass your diffs to someone who knows the software you're patching, and they review them before committing them. That's to make sure that the person who's sending you a fix for some particular piece of hardware doesn't break support for something else that they can't test. Obviously this wouldn't apply in Theo's case, but it's hardly an uncommon situation.
think his "utter asshole" reputation is not accurate
I don't think I'd add "utter" to that description, but from what he's written (that I've read), he certainly seems like an abrasive personality with a short fuse. However, he is the lead of a successful software project, so he must have something going for him...
Problems? I wasn't aware that NetBSD really had any problems, per se, it just didn't have all the features that many other OSes had. I mean, if you're picking NetBSD for its network stack, chances are its primitive support for sleep mode in notebooks is pretty much a non-issue.
If you'd read the article, you'd have seen that Linux actually existed first. NetBSD arose because Linux was still too much of a toy at that point. TFA makes a good point for why Linux ultimately became more "successful" (as far as being a dynamic, viable system, I don't know how they compare installed-base-wise).
It's a shame, I really like NetBSD (even if a lot of their "supported architectures" just boot the kernal and may not really run anything useful), and I hate to see them in this state. Oh well, that's the beauty of OSS: if things really get bad, somebody will just fork the latest codebase and take it in a better direction (c.f. OpenBSD, DragonflyBSD....)
I imagine the reason for the restricted frequency range is due to the difficulty of designing an antenna that is reasonably efficient that covers that much spectrum. I know that I have separate antennas for 80M, 40M, 20M and 2M/70cm, and I know that it's tricky to design and build antennas for the higher bands (a full-wave dipole for 3.8GHz would be ~1.5", and dipoles aren't especially suited for higher-frequency work). Sure, you could use a Pringles can, but you'd need several to cover the range of frequencies for true UWB. I don't know of any trapped antenna designs in the GHz range, but that might be something to try...
Just make sure you know it's there...
on
16GB Flash USB Dongle
·
· Score: 2, Interesting
I backed up my files to my flash drive, and then left it in the USB port and proceeded to upgrade Solaris. It kept complaining that my primary boot drive was only 512M and only had 15M free. I thought something was fuxx0r3d with my partition table, until I saw what the mount point was... At least now I know my box can boot from USB, that may come in handy, given a sufficiently sad set of circumstances.
That's part of his point, in order to accept the OSS he (or someone on his team) would have to write the design specs, then validate the code against them. Which would basically take as much time and effort as writing it from scratch themselves.
FWIW, I have written FDA-approved software (LIMS), and I know I wouldn't (and didn't) use any third-party stuff in any of my systems if I could write it myself. It takes more time, but we were known as miracle workers at my company, because we actually delivered working systems within a reasonable facsimile of our target date. Pharmaceutical R&D development teams are notorious for eating millions of dollars and spending years chasing state-of-the-art systems that they've convinced some department head they need, only to wind up scrapping the project because it doesn't work or isn't performant. It's like Defense Department work, only with company money. Of course, these are pharmaceutical companies, so it's not like they miss the odd thirty or fifty million...
Sony is counting on the PS3 turning around the company's fortunes
So Sony, who owns motion-picture studios, music labels, and manufactures sh!tloads of consumer electronics gear besides the PlayStation line, is depending on the PS3 to reverse its downward slide? Anybody know what percentage of Sony's profits come from PS sales? Can it even be broken down that far, or is it just lumped in with the other consumer electronics gear?
Sorry, I have a hard time believing that just one product can make or break a company Sony's size.
I definitely think that Linux (in its current state) is for tweakers and fiddlers
Is it still that bad? I used Red Hat 5.1 for awhile, then switched to Debian Potato (man, I hated that configuration tool, dselect?), then got tired of downloading utilities only to find they were some high-school senior's CS project and only worked with a specific (obsolete) version of some common library that I couldn't install because it would conflict with the current version which I needed for everything else. So I moved on to NetBSD and never looked back. I've been thinking about loading Ubuntu on one of my mothballed machines, but maybe I won't bother...
Apple probably won't support you if you use hardware other than what they sell
I wouldn't be surprised if Apple released some kind of "generic" version of OSX server that could run on white boxes. You wouldn't have the GUI and all the iLife software, but it'd be the core of the OS, possibly enhanced for better I/O throughput. If they can leverage iPods to get Macs into people's homes, maybe they can leverage OSX on the server to get Macs into businesses (esp. if they bundle their dynamic cluster management software -- which runs on anything, but has some fantastic features that only work on a Mac).
This seems like a reasonable limit to me. The last time I installed Solaris it wanted to give me ~800M (on a machine with 1G of memory), I overrode it to take a gig. Of course, under Solaris/tmp is mapped to swap space, so it's usually a good idea to give yourself a little more room anyway. OTOH, even debugging stuff in NetBeans while running Sun's app server, I've never eaten more than half a meg of swap, so it looks like Solaris at least has a pretty good swapper. I do recall the only time I ever had NetBSD crash on me, I was rebuilding the kernal and passed "-j" as an argument to make(1) (which tells it to create as many subprocesses as it wants to process things in parallel). Ran out of swap and everything went dark....
When I lived in Oregon we would get coupons from the power company for free CFLs. Two or three at a time. When I mentioned that I liked them, I got probably a couple of dozen from cow-orkers who said they already replaced all the bulbs in their house they could. The only problem I had with them was the fact that they looked odd in my ceiling cans. I did buy a couple in the mid-90s when they first became popular, and I wasn't impressed by the color or the brightness, but the ones I got in Oregon were fine.
Heh, I was in the meeting the day my boss suggested we change the name of the about-to-be-launched Agricultural Info-Data Service, due to the "unfortunate connotations" the acronym might suggest...
I wonder if there will eventually be a dependent subculture built around grids. Companies who sell pre-written (and customizable) solutions that require a grid to run. Companies might buy/license them and run them quarterly, to (say) update/analyze their data warehouse. Or if somebody develops a "killer app" for grids, then they might set up their own and rent it as a distributed computing solution. Maybe develop some kind of adapter for Mathematica, let people run their huge econ simulations, then truck their results away to slice & dice.
It's as though the judge is saying "Well done son! Here, have a cookie!"
In other news, it was revealed that the judge in this case is also a former employee of the firm, and left under a cloud following a boisterous office party...
I hope the players can in-game construct an appropriate response to the guy.
Um, what's to prevent this guy from just leaving the game? His character goes into limbo, with all the money, and that's the end of it. Maybe he signs up again under a different name, maybe he just moves on to WoW or whatever.
I think that's the biggest problem with this kind of scam -- at least IRL, you can't just leave (well, not without a fair amount of effort, anyway).
I assumed that the button would send user info for both sides, as well as a transcript of the conversation, so that the officer/investigator could evaluate the case and make a good preliminary determination. You wouldn't be able to just "press the button" and report someone by their e-mail address.
The way I'd work it would be to have the aforementioned info sent, then have the police quickly follow up with a quick e-mail message to the accusing party with something like "If you really meant to send this, click here to confirm" (like signing up to some mailing lists). Only after the sender confirmed the submission would it actually be delivered to some appropriate inbox.
Might also be a good idea to send e-mail to the accused, too, to let them know what's coming down.
Dunno, I always thought the Aquos line sucked ass -- muddy colors and dim output. I bought a 42" Panasonic plasma, mounted it on the wall and have enjoyed the hell out of it (coupled with a nice receiver and appropriate speakers). I also just bought a 32" Polaroid LCD and there's no comparison -- the contrast ratios and color levels are nowhere near the same. Of course, the LCD cost about half what the plasma did (I wouldn't have bought it otherwise), but it's easily half the set.
<rant>What's the big deal about plasma "burn-in"? You turn on the set, watch whatever, then turn it off. What are you people doing with your sets that would burn them in? Is this an issue with video games? I know my plasma had some kind of new feature that was supposed to "practically eliminate" (for whatever that's worth) burn-in. I can understand how you might have an issue in a commercial environment if you just display a logo or play a looped recording for sixteen or eighteen hours a day, but what's the big deal on a home set? </rant>
As I understand it, it was partly the heat, and partly the inability to physically get enough power to the chip circuitry. As process shrinks made the individual circuits smaller, the individual traces became less able to carry the power required by the 20-stage+ pipeline. Making the traces bigger made it harder to lay out the chip, and negated some of the advantages of the process shrink.
There's an interesting explanation of the whole situation here.
The biggest mistake Intel made was assuming... that raw Ghz would still be the biggest influence on processing speed
I doubt they made that mistake, I think it's more likely they saw that raw GHz was the biggest influence on purchasing decisions. Why do you think AMD revised their processor numbering scheme to emphasize "apparent" processor speed? Do you honestly think your average computer buyer would pick a dual-core 2.2GHz processor over a 3GHz single-core, even though the dual-core may have 40% better[*] throughput?
I agree that NetBurst failed to scale to the degree that Intel was hoping, but I'm sure marketing was one of the largest driving forces behind driving processor speeds at the expense of almost everything else.
Slashdot Mirror
I've been on a couple of projects where there was a lone woman (either developer or QA), and they were always treated as just another team member. OTOH, both times we were all on new teams, so there was never that awkward "new kid on the block" phase.
You haven't contributed to many OSS projects, have you? That's actually the way most contributions from new developers get made: you pass your diffs to someone who knows the software you're patching, and they review them before committing them. That's to make sure that the person who's sending you a fix for some particular piece of hardware doesn't break support for something else that they can't test. Obviously this wouldn't apply in Theo's case, but it's hardly an uncommon situation.
I don't think I'd add "utter" to that description, but from what he's written (that I've read), he certainly seems like an abrasive personality with a short fuse. However, he is the lead of a successful software project, so he must have something going for him...
Problems? I wasn't aware that NetBSD really had any problems, per se, it just didn't have all the features that many other OSes had. I mean, if you're picking NetBSD for its network stack, chances are its primitive support for sleep mode in notebooks is pretty much a non-issue.
If you'd read the article, you'd have seen that Linux actually existed first. NetBSD arose because Linux was still too much of a toy at that point. TFA makes a good point for why Linux ultimately became more "successful" (as far as being a dynamic, viable system, I don't know how they compare installed-base-wise).
It's a shame, I really like NetBSD (even if a lot of their "supported architectures" just boot the kernal and may not really run anything useful), and I hate to see them in this state. Oh well, that's the beauty of OSS: if things really get bad, somebody will just fork the latest codebase and take it in a better direction (c.f. OpenBSD, DragonflyBSD....)
I imagine the reason for the restricted frequency range is due to the difficulty of designing an antenna that is reasonably efficient that covers that much spectrum. I know that I have separate antennas for 80M, 40M, 20M and 2M/70cm, and I know that it's tricky to design and build antennas for the higher bands (a full-wave dipole for 3.8GHz would be ~1.5", and dipoles aren't especially suited for higher-frequency work). Sure, you could use a Pringles can, but you'd need several to cover the range of frequencies for true UWB. I don't know of any trapped antenna designs in the GHz range, but that might be something to try...
I backed up my files to my flash drive, and then left it in the USB port and proceeded to upgrade Solaris. It kept complaining that my primary boot drive was only 512M and only had 15M free. I thought something was fuxx0r3d with my partition table, until I saw what the mount point was... At least now I know my box can boot from USB, that may come in handy, given a sufficiently sad set of circumstances.
That's part of his point, in order to accept the OSS he (or someone on his team) would have to write the design specs, then validate the code against them. Which would basically take as much time and effort as writing it from scratch themselves.
FWIW, I have written FDA-approved software (LIMS), and I know I wouldn't (and didn't) use any third-party stuff in any of my systems if I could write it myself. It takes more time, but we were known as miracle workers at my company, because we actually delivered working systems within a reasonable facsimile of our target date. Pharmaceutical R&D development teams are notorious for eating millions of dollars and spending years chasing state-of-the-art systems that they've convinced some department head they need, only to wind up scrapping the project because it doesn't work or isn't performant. It's like Defense Department work, only with company money. Of course, these are pharmaceutical companies, so it's not like they miss the odd thirty or fifty million...
Sorry, I have a hard time believing that just one product can make or break a company Sony's size.
Welcome to the 21st century. Watch your head. Mind the gap.
This seems like a reasonable limit to me. The last time I installed Solaris it wanted to give me ~800M (on a machine with 1G of memory), I overrode it to take a gig. Of course, under Solaris /tmp is mapped to swap space, so it's usually a good idea to give yourself a little more room anyway. OTOH, even debugging stuff in NetBeans while running Sun's app server, I've never eaten more than half a meg of swap, so it looks like Solaris at least has a pretty good swapper. I do recall the only time I ever had NetBSD crash on me, I was rebuilding the kernal and passed "-j" as an argument to make(1) (which tells it to create as many subprocesses as it wants to process things in parallel). Ran out of swap and everything went dark....
When I lived in Oregon we would get coupons from the power company for free CFLs. Two or three at a time. When I mentioned that I liked them, I got probably a couple of dozen from cow-orkers who said they already replaced all the bulbs in their house they could. The only problem I had with them was the fact that they looked odd in my ceiling cans. I did buy a couple in the mid-90s when they first became popular, and I wasn't impressed by the color or the brightness, but the ones I got in Oregon were fine.
He's no fun, he fell right over!
I wonder if there will eventually be a dependent subculture built around grids. Companies who sell pre-written (and customizable) solutions that require a grid to run. Companies might buy/license them and run them quarterly, to (say) update/analyze their data warehouse. Or if somebody develops a "killer app" for grids, then they might set up their own and rent it as a distributed computing solution. Maybe develop some kind of adapter for Mathematica, let people run their huge econ simulations, then truck their results away to slice & dice.
Well, duh. You've got the source, debug it!
Kids these days...
I think that's the biggest problem with this kind of scam -- at least IRL, you can't just leave (well, not without a fair amount of effort, anyway).
I assumed that the button would send user info for both sides, as well as a transcript of the conversation, so that the officer/investigator could evaluate the case and make a good preliminary determination. You wouldn't be able to just "press the button" and report someone by their e-mail address.
The way I'd work it would be to have the aforementioned info sent, then have the police quickly follow up with a quick e-mail message to the accusing party with something like "If you really meant to send this, click here to confirm" (like signing up to some mailing lists). Only after the sender confirmed the submission would it actually be delivered to some appropriate inbox.
Might also be a good idea to send e-mail to the accused, too, to let them know what's coming down.
BSD was hardly a major operating system at that time. A major research system, maybe.
Dunno, I always thought the Aquos line sucked ass -- muddy colors and dim output. I bought a 42" Panasonic plasma, mounted it on the wall and have enjoyed the hell out of it (coupled with a nice receiver and appropriate speakers). I also just bought a 32" Polaroid LCD and there's no comparison -- the contrast ratios and color levels are nowhere near the same. Of course, the LCD cost about half what the plasma did (I wouldn't have bought it otherwise), but it's easily half the set.
<rant>What's the big deal about plasma "burn-in"? You turn on the set, watch whatever, then turn it off. What are you people doing with your sets that would burn them in? Is this an issue with video games? I know my plasma had some kind of new feature that was supposed to "practically eliminate" (for whatever that's worth) burn-in. I can understand how you might have an issue in a commercial environment if you just display a logo or play a looped recording for sixteen or eighteen hours a day, but what's the big deal on a home set? </rant>
As I understand it, it was partly the heat, and partly the inability to physically get enough power to the chip circuitry. As process shrinks made the individual circuits smaller, the individual traces became less able to carry the power required by the 20-stage+ pipeline. Making the traces bigger made it harder to lay out the chip, and negated some of the advantages of the process shrink.
There's an interesting explanation of the whole situation here.
I agree that NetBurst failed to scale to the degree that Intel was hoping, but I'm sure marketing was one of the largest driving forces behind driving processor speeds at the expense of almost everything else.
* Numbers courtesy My Ass Research Associates