Oddly enough, however, Homeland Security is going to be one of the places that you can be assured of hiring geeks grown in the USA. Why? The government isn't too keen on foreign nationals running even small parts of our national security, much less writing code that will be used for their systems.
Beyond working directly for "the man", the government, including homeland security, is generally happy to hand out wads of cash for interesting research that they think can provide some sort of value to them, even if indirectly. Even if our local corporations are concerned about maximizing their bank balances at the cost of the geeks there are other opportunities.
Personally, I think US corporations need to remove their collective heads from their behinds and focus on a little bit of long term strategy. This new crop of get rich quick CEOs is getting on my nerves, opposed to the let's dominate everything in sight sort I've been told existed in days gone by. Homeland security may be frightening off foreign students, but US CEOs must be frightening off students who would stay in the US for a job, domestic and foreign alike.
We don't use a free and open DNS system because it would encourage others to do the same, just not our free and open DNS system.
You think Microsoft couldn't hose their TCP/IP code to default to a certain DNS server when DHCP is used, defaulting to the real DNS server when the name doesn't resolve? Oops, suddenly linux.org tells you the communist evil of linux. Only half jest, as it'd get MS in deeper water, but I've made my point.
We're talking a whole slew of fractures if people start getting the idea to make many "root" servers. ICANN may not be the greatest thing in the world, but they maintain consistancy. That consistancy needs to be around for the net to function right for everybody. I for one don't want to have to register my domain with a dozen places to make sure that it resolves properly everywhere.
DNS traffic is pretty small, so don't think a bunch of servers wouldn't pop up either. Wouldn't it be great if ICANN had 50% of resolutions and there were 10 others competing with 5% of resolutions each, none of them consistant with each other and all dependent on your ISP? No, I didn't think so either. Oh, and imagine the joy of other people getting your email when somebody has squatted your domain in one of the dozen inconsistant "root" servers.
The only place I've really had a chance to play with SCSI disks is really in the workplace, and I have to say that my experience with them is much better than those I've had with IDE drives in the workplace. Sure, we've had SCSI drives go down, but there are plenty of nice SCSI RAID controllers that do hot spares and multiple RAID 5 arrays on the same controller and the like. I've had nothing but issues when using IDE in a server setting where the disk gets pounded, and then it tends to be RAID 1 without a hot spare when IDE is concerned, which just left me nervous when one of those drives would fail. In my non-critical home, IDE does the job fine, as I tend to bang on the drives for brief periods, not constantly.
It also took the SCSI drives a lot longer to fail it seemed, and that they often come with at least a 5 year warranty is nice. As far as a rendering cluster is concerned, I'd assume that the nodes responsible for distributing frames to be rendered would have plenty of redundancies in place, but the boxes used to do the dirty work would not. Rendering is mostly CPU and RAM anyway.
It's like you say, a handful of boxes can disappear and the work still goes on. That's the nice thing about movie rendering, the task is so explicitly and locally parallel because each machine can take a set of frames and render without having to share data. Losing one or two for a while wouldn't hurt the operation, but losing one or two regularly would be a real problem for the staff. After a year of constant use I'd think most problems would start occuring, as that's when basic warranties on most computers start running out, Apple's included. And with some 6000 computers, a lowly 1% chance of some part failing means 60 computers probably will. Yuck.
Sorry if I came across as making it out to be the holy grail, but I do think power is something that is oftentimes overlooked as a factor in computing.
My comment about worrying about components failing was more a comment on the ever reducing warranties on parts. 1 year warranties on new off the shelf hard drives is along those lines, though high end scsi isn't so likely to fail. For something that's just rendering frames the disk could probably be normal ATA though. With the mean failure rates obviously being lowered on commodoty parts, swapping out parts after a year would probably prevent a good deal of little failures, either PC or Mac. With thousands of computers they're bound to happen too.
I said that it was another advantage, not the sole reason to upgrade. Anyway, assume that each machine draws 300W continuously for a month. I don't have the specifics on Pixar's machines, but I believe I read the farm used for LotR was 6000 machines strong. (Odd that it isn't a power of 2 though)
So 300 W * 24 h * 31 days * 6000 computers / 1000 kW/W = 1,339,200 kWh per month for all those machines. Assuming 7 cents to the kWh is $93,744 in power alone each month. That's possibly undershooting, as I don't know what the cost per kW is over lines that deliver some 40 thousand kW each hour.
Now, tack on the costs of cooling this beast of a farm, the monthly cost just to have lines it takes to deliver over 40 thousand kW each hour and that total just goes up. It certainly doesn't break the bank of a company producing huge films, but it is an advantage to reduce that number.
I never claimed that power was the only or the best reason to upgrade. I'm just saying that this is a cost that is often overlooked, and that the savings could offset some of the upgrade costs. When you have a good bulk purchase program as Pixar probably can get through Apple it's even better. I figure they can sell the used machines at a decent rate if they're but a year old, or at least donate them and write them off.
Heck, there are probably dozens of reasons to upgrade often in that business. Not having to worry about parts that most often fail after a year, space concerns when increasing computing power, being seen as a pioneer because you have the latest and greatest, more time to test render, etc. Power, however, is still going to be among the list.
There's another advantage of faster speed, beyond just being able to crank out more of the same quality frames per minute, or cranking out higher quality frames in the same time. Power cost. If two machines use the same amount of electricity while one renders faster, well, the amount of money you spend on power is reduced. If this takes a month off the total render time it's great, though the movie probably won't be released earlier. However it is a month worth of a massive cluster's power bill that's being saved.
I imagine that on a large enough scale operation, the cost to upgrade anually is decently offset by the power savings from not running the machines as long for the same output. I'm sure the remaining cost is easily made up for in the value of earlier release. Or along the route of higher quality frames, the same amount of power cost plus more in depth graphics is valuable to be seen as the pioneers in the field, plus having more visually appealing movies.
It is probable also very much what you're saying that hardware is a drop in the bucket compared to the amount they're making.
For those who are using older hardware which has your recommended base of 64 MB of RAM, there will always be the option to use lightweight window managers. Ice, tvwm2, etc. Why is it that there are people with low end hardware that complain about not having all the bells and whistles that can be had on high end kit while completely ignoring very usable software that will run on their machines without sacrificing any real important features?
It's kinda like using an S3 Virge 8MB PCI video card and complaining about low Quake 3 framerates. Either turn down the textures and details and play the game at low res, upgrade, or play Doom. There's nothing wrong with Doom, just like there's nothing wrong with lightweight window managers.
I use several machines that cross the scale from Pentium Pro to Pentium 4 and Athlon XPs. My main desktop, a Pentium 4 with 1 GB of RAM runs Gnome 2 great, ran Gnome 1.4 great as well. Gnome 2 is actually faster than 1.4, and has been anywhere I've used it, regardless of memory size. At any rate, I want them to feel free to use resources as long as it doesn't bloat to the point where I have issues running a slew of programs at once. So far it hasn't and I'm a big fan of the improvements that keep going into Gnome. I also know better than to try putting Gnome on the Pentium Pro. It just isn't a great idea.
Beyond that, even most Pentium 2 systems have provisions for 256-512 MB of memory. The Pentium 2 400 I have at work ran Gnome 2 fine on 128 MB under conservative usage, ie not having 2 dozen programs open at once. It's kinda common sense that the machine will swap at that point. I ended up with more RAM, but that was because I was running RAM happy experiments. Gnome wasn't that big of a factor.
And RAM isn't exactly expensive either. I understand not wanting to pay any more on hardware to run programs, but again, if you want lightweight there are many, many other programs out there that are. They don't give you the moon on a stick, but are very reasonable alternatives for boxen that have small quantities of RAM.
As I found out a few days ago just perusing through my CD binder, Encarta was indeed bundled with my Dell Laptop as part of the Works Suite. I've never installed it, but would consider doing so if I had disk space to spare, as I'm not always near a wireless hotspot. For the size though, it isn't worth the use I'd get out of it.
When I saw it sitting there in my binder though, the first thought I had was, "Gee, don't see many encyclopedias around these days..."
Well, I don't watch TV. I think I've watched the news once in the past 3 months or so. I watch movies on a semi-regular basis, but that's at my discretion of movie and time. Most anything I watch is anime, and that tends not to be quite a bit less than even 10% of my free time.
I find strangers that just walk up to me on the street and start talking to be extremely creepy. Yes, I'm an introvert, but I have plenty of friends and hold conversations and even arguements with people I know without issue. What in the world posseses somebody to just strike up a conversation with me I'll never know.
I wear headphones when I'm out walking alone very often because I love listening to music and am not planning to interact with anybody. That doesn't mean I won't take them off if I see interaction coming up. I did so just last week to help somebody take a picture of themself against the Minneapolis skyline. No biggie, and it wasn't somebody randomly wanting to hear my life story or tell me theirs. The only time it has ever stopped interaction that I knew of and I didn't take them off is the religious nuts that stop everybody they think will listen. But I've stopped and played their games and debated philosophy with them too. At least I know why they're talking to random people.
I smile at people as I walk past them, and often they smile back. It's polite and unobtrusive. I'll never just randomly start a conversation though. People that do just weird me out way too much.
One of the reasons is that there are varying lengths to a variety of words in Japanese and English. The dialogue may take 10 seconds to say in Japanese and boil down to about five English words (minor exageration). When this happens, you have to fluff up the English a bit or it looks like a bad Godzilla dub. We're talking no sound for a couple seconds and the mouth is still moving.
A subtitling is under no such pressure. If the subtitle takes its sweet time to say the same thing, or puts the sentence into the five words it really means, well, it does. You read, you comprehend, you get sound with the mouth moving, even if it isn't perfect. (You don't think the Japanese always lines up perfect, do you?)
There's also another thing that fansubbing does at least. That is to use a few Japanese words (well, their romanizations) in the sub itself. The average watcher is generally pretty clueless to it, so I don't think it'd be a mainstream (read: Disney) thing, but it adds a lot when these little things are carried over because they have a different feel to them than the English translations tend to. Especially the suffixes used with names, sensei, sama, chan, kun, etc.
When these are translated to what they mean, like little Yahiko in the Kenshin DVD subs (can't stand the English voices) instead of Yahiko-chan like in the fansubs it makes my skin crawl. Little Yahiko is not something that anybody'd think to say in English except maybe in the instance of "poor little Yahiko" which would have waaay been better in the Kenshin instances because chan is used as mocking Yahiko. Anywhere else and you don't really think to call a child little, so "Thank you little Chihiro" is right out, where "Arigato Chihiro-chan" is, well, proper (even if I may have killed the spelling). So it invariably gets dropped in the dubbing process. This leads to a different problem, you lose the information of the relation of two characters by not using these methods of address. Sama and sensei are terms of respect, and to call Yahiko by the term Yahiko-sama means you have respect for them. Instead it's just lost, or translated into a Mr. Yahiko, which in a lot of cases sounds dumb and doesn't imply the same respect.
Just my two cents on the whole dubbing process. And before anybody thinks I'm being too harsh, let me say I have a lot of respect for the people that dub in both the original Japanese and the English releases. I've been working on a fandub with some friends (bad idea, but fun to try) and, while it's coming along, it is incredibly time consuming and difficult to get the voices to sound consistantly in character, let alone capture what's being said in approximately the proper length of time and sound like you mean it. I imagine there are similar problems getting from Japanese script to English subs, just to make the lines sound right when reading, but I think there's less pressure to change things to fit the length of time that mouths are moving and thus the sub can be more true to the original Japanese.
Well, guess it's about time to dust off the old talkie CD version of Sam and Max hit the road. Of course, once you've gone through the game it's kinda hard to forget all the wacky hijinks and the general method of solving some of the most interesting and bizarrely strung together puzzles I've ever seen.
Guess I'll get around to that letter crying to Lucasarts too. I was looking forward to this game. However, if anybody hasn't read a Sam and Max comic book I suggest picking one up. It might ease the pain a little bit.
Excellent point, though I suppose they could just mandate a certain font at a certain width with certain mandatory kerning options and standardized margins and... yeah, word count is probably just easier.
Damn, and I'm in college still. I should've been able to guess that one. Guess grad school is out of the realm where many people feel the need to use 30 point wide fixed width fonts to demonstrate their incompetence. Plenty of ways to do that with content.
Yes, I've heard this as well. Word I think does not count the words in the footers and footnotes toward the total. The result: several fines being given for lawyers that went beyond the number of words they were allowed.
Just what I've heard, but it'd make a sort of logical sense as to why to stay away from Word if you're in law. Of course, I'm not sure why there'd be a hard word maximum and not just a page maximum. But there's a lot of quirks in law I'm not so sure of.
I don't think that children, especially 16 or under should be given enough privacy to allow them to secretly purchase a game and play it. Maybe they can get away with doing that kind of crap at a friend's house, but there should be constant monitoring by parents of what's going on at other homes as well as theirs. Sure, kids hate it, I know I wasn't keen on the idea, but it helps in the long run.
As for why most young people are involved in homicides, I remember doing the research for some Fark topic and finding that there's way more young people involved in killing or being killed than you'd expect, but most of it is gang related. Usually these crimes are perpetuated in the neighborhoods that are least likely to have a bunch of pasty white suburban children (excuse my stereotyping) imitating GTA. They've been living out the real thing without ever playing for a while.
When the pasty white suburban kid does go nuts and kill though it seems almost certain to be because the parents could not keep tabs on their child through the developmental stages. This is probably because the kids got themselves confused about real and fake when they were impressionable, completely at the fault of the parents. The same is really true that if parents could exercise a little more control even in bad neighborhoods that there would likely be less violence. I'm just going to be realistic about it and say that the gang neighborhoods are in a state where it's way too easy to propogate violent tendencies and the best move a parent could make is to leave if they can.
Some people may be sociopaths, but I doubt that all that many really are. It doesn't seem like a trait that would be prone to stick with society too well and, well, frankly isn't going to be expressed or not expressed because of a video game.
I'd really just bet that aside from a few freak genetic mishaps, most of the problems of violence in our society are primarily due to poor parenting, honestly.
Sure, a person could be either amoral or a moron because they're influenced to kill by a video game. What made that person that way? I'd say it's likely that their parents made a few too many bad choices, didn't pay enough attention, or thought that children raise themselves.
It's high time victims in this country start suing parents of underage perpetrators of violent crime. Somebody's 16 year old kid tries to mow you down with their car? You should be able to sue the hell out of their parents for negligence of some sort. That'd teach parents to stop shifting blame for their being crappy moral compasses. Maybe then they'd start to pay attention to what is raising their children when they aren't. Be it video games, movies, beer commercials, or even the 10 o'clock news.
Yeah, I think that on an impressionable mind, especially that of somebody under 8 years of age, any sort of violent media could easily be misinterpreted. That's why there's parents. Adult entertainment is labeled as adult entertainment because it's for adults. If an adult cannot handle it then it's probably because of something they were taught as a child.
Does he mean that 40 people, instead of 20 people, will be killed as a result of someone imitating the game in real life?
No, I think he means that sane people will have more fun playing the game than they might expect out of a sequel which probably will have only incremental improvements to its engine.
Expectations about stupid, homicidal people being stupid and homicidal are already high, regardless of their playing a video game or desire to have a scapegoat for their stupid and homicidal actions. As such, I'm pretty sure he was mentioning just the game aspect.
While your fame may not necessarily spread outside of geekdom, and is likely to not even spread within it, there is still benefit to being one of the authors. If your name shows up in an about box or somewhere accessible, you can put in your resume or show in your interview that you're partly responsible for the code. If it's a popular application in the geek community then hopefully one of the people doing the interview will have heard of it if not you.
Beyond that, I think it provides an excellent answer to the "What can you bring to the company?" question. Being open they can inspect what the software you've codeveloped is like and see if they like it. It's win-win because you get to show off and the company gets to see the quality of the developer they're about to hire. Try to get that sort of perk from your last proprietary company.
Finally, open source projects provide something in a poor student's free time that they probably aren't apt to get during their college career. That's experience. I wrote a three hundred line program on my own for a class doesn't sound nearly as good as being responsible for even a few hundred lines in a much larger project, even if nobody really knows about either. And again, if it's a popular project then so much the better. What's the chances of working on a big project with no experience during college?
And these are just the personal benefits of having written some of the code. While working on the Linux kernel may not pay the rent, paying hundreds of dollars every time you want (or even need if you're a developer) to get a new version of some particular piece of software doesn't help pay it either. Think for a moment how much of a person's salary could go towards paying for the professional version of a proprietary operating system, office suite, database, and development studio. $2,500 down and about half of that every time there's an update, which seems to be every 2-3 years. Of course, last I checked that doesn't provide anything that would allow you to run your own webserver to try developing web applications, nor does it provide you with a source code repository for version tracking, nor could you run an email server if you wanted to try your hand at administering one of those. List keeps going.
There's a lot of benefit to everybody when there is free software at hand. There's plenty of software left to be written and companies will pay for that. But there's also a lot of software that already exists and isn't going to make anybody except the big companies money when it's sold while the company pays its workers to make more software. It is also likely cost people interested in the technical side of it more, since you can't just learn about it for free.
Oh certainly, I wouldn't disagree with that at all. It just seems that all too often people look at stress as being absolutely bad.
I actually like having a little stress in my life now and again. Things could be awfully dull without it. And the best part is triumphing over those stressful challenges and feeling good about it.
I couldn't imagine being stressed to the point of heart attack though. Maybe I manage it well, but it just needs to be set aside if it gets that bad. Failure isn't the worst thing in this case.
You can cause yourself stress as well by not working to your own personal feeling of what your level of quality should be. If you care a lot about how the mail server you put together is working and it goes down, it can be a stressful event simply because you were expecting it to stay up and you want to get to the cause of what made it crash.
Apart from that, everybody keeps looking at stress as though it's a bad thing. I know some of my best work is done under the heavy pressure of stress. While a lot of stress certainly wouldn't be a healthy level for me to maintain, a bit of stress, even really intense stress, can be good for you and keep you from being complacent. I'd hate to lead a completely stress free life.
That's what I seem to remember as well. If I were Intel I'd probably just play it down along the lines of:
"Well, most people still have no use for these instructions, and we could have developed a set of 64 bit extensions to our own architecture on a whim, but in the spirit of preventing confusion about which extensions are in use on our x86 instruction set and to allow for easier forward progress when people do need 64 bit computing, we'll just implement what AMD threw together. You know, just because it's there, not because it's difficult. Why waste money on developing diverging 'standards'?"
That'd be my spin anyway, regardless of how wrong it may be.
No, and those who adopt early will make some difference to the market share. Lots of businesses, however, don't have the money to rollout hundreds of new processors very often. Heck, the last company I worked for (100s of employees, millions in monthly sales) is still on Pentium 3 (some Xeon) machines for servers, except for a single database box which needed more.
While I think the Opteron is a great choice, and kudos to your company for getting the best of what they can right now, many companies are also either blindly loyal to Intel or don't want to adopt anything too "new" too quickly. The fact that a few linux distros and now some beta software from MS will take full advantage of the Opteron will probably have an effect too, despite the Opteron running 32-bit code very nicely. It's the uninformed managers that tend to make those decisions.
Again, congrats to your company which decided to jump on already. I'm still of the opinion that there will be plenty of 64-bit sales for Intel and this is no more than a minor scrape, completely detached from a favor for either company.
If they're willing to go the x86 with 64 bit route they don't need to worry about being years behind. Itanium is a bit of a mess, sure, but as far as I remember Intel is allowed to use whatever improvements AMD makes to the x86 architecture free of charge.
Also, it's been stated elsewhere that the Prescott cores already have the extensions, they just aren't enabled. So there's apparently no problem getting the extensions working on the chips, in the worst case they don't have to develop any of the extensions, and AMD has blown a bunch of cash raising interest in 64 bit processors.
I think Intel probably isn't going to be hurting so bad here.
No, you're right in that Sun definately isn't winning the desktop share away from Microsoft. CDE is almost certainly my guess as to one of the reasons why, and I wouldn't dream of making my Grandma use it.;)
It seemed for a while though, that Sun was at least trying to make Solaris a desktop friendly OS. They've been putting work into their office suite and have adopted Gnome as a WM in addition to CDE. Though from what I hear now, they're trying to make their Java Desktop which runs Linux their main desktop OS. Regardless, Sun is trying to take desktop share away from MS and is still around. Which is more than we can say about Be.
Slashdot Mirror
Beyond working directly for "the man", the government, including homeland security, is generally happy to hand out wads of cash for interesting research that they think can provide some sort of value to them, even if indirectly. Even if our local corporations are concerned about maximizing their bank balances at the cost of the geeks there are other opportunities.
Personally, I think US corporations need to remove their collective heads from their behinds and focus on a little bit of long term strategy. This new crop of get rich quick CEOs is getting on my nerves, opposed to the let's dominate everything in sight sort I've been told existed in days gone by. Homeland security may be frightening off foreign students, but US CEOs must be frightening off students who would stay in the US for a job, domestic and foreign alike.
You think Microsoft couldn't hose their TCP/IP code to default to a certain DNS server when DHCP is used, defaulting to the real DNS server when the name doesn't resolve? Oops, suddenly linux.org tells you the communist evil of linux. Only half jest, as it'd get MS in deeper water, but I've made my point.
We're talking a whole slew of fractures if people start getting the idea to make many "root" servers. ICANN may not be the greatest thing in the world, but they maintain consistancy. That consistancy needs to be around for the net to function right for everybody. I for one don't want to have to register my domain with a dozen places to make sure that it resolves properly everywhere.
DNS traffic is pretty small, so don't think a bunch of servers wouldn't pop up either. Wouldn't it be great if ICANN had 50% of resolutions and there were 10 others competing with 5% of resolutions each, none of them consistant with each other and all dependent on your ISP? No, I didn't think so either. Oh, and imagine the joy of other people getting your email when somebody has squatted your domain in one of the dozen inconsistant "root" servers.
It also took the SCSI drives a lot longer to fail it seemed, and that they often come with at least a 5 year warranty is nice. As far as a rendering cluster is concerned, I'd assume that the nodes responsible for distributing frames to be rendered would have plenty of redundancies in place, but the boxes used to do the dirty work would not. Rendering is mostly CPU and RAM anyway.
It's like you say, a handful of boxes can disappear and the work still goes on. That's the nice thing about movie rendering, the task is so explicitly and locally parallel because each machine can take a set of frames and render without having to share data. Losing one or two for a while wouldn't hurt the operation, but losing one or two regularly would be a real problem for the staff. After a year of constant use I'd think most problems would start occuring, as that's when basic warranties on most computers start running out, Apple's included. And with some 6000 computers, a lowly 1% chance of some part failing means 60 computers probably will. Yuck.
My comment about worrying about components failing was more a comment on the ever reducing warranties on parts. 1 year warranties on new off the shelf hard drives is along those lines, though high end scsi isn't so likely to fail. For something that's just rendering frames the disk could probably be normal ATA though. With the mean failure rates obviously being lowered on commodoty parts, swapping out parts after a year would probably prevent a good deal of little failures, either PC or Mac. With thousands of computers they're bound to happen too.
So 300 W * 24 h * 31 days * 6000 computers / 1000 kW/W = 1,339,200 kWh per month for all those machines. Assuming 7 cents to the kWh is $93,744 in power alone each month. That's possibly undershooting, as I don't know what the cost per kW is over lines that deliver some 40 thousand kW each hour.
Now, tack on the costs of cooling this beast of a farm, the monthly cost just to have lines it takes to deliver over 40 thousand kW each hour and that total just goes up. It certainly doesn't break the bank of a company producing huge films, but it is an advantage to reduce that number.
I never claimed that power was the only or the best reason to upgrade. I'm just saying that this is a cost that is often overlooked, and that the savings could offset some of the upgrade costs. When you have a good bulk purchase program as Pixar probably can get through Apple it's even better. I figure they can sell the used machines at a decent rate if they're but a year old, or at least donate them and write them off.
Heck, there are probably dozens of reasons to upgrade often in that business. Not having to worry about parts that most often fail after a year, space concerns when increasing computing power, being seen as a pioneer because you have the latest and greatest, more time to test render, etc. Power, however, is still going to be among the list.
I imagine that on a large enough scale operation, the cost to upgrade anually is decently offset by the power savings from not running the machines as long for the same output. I'm sure the remaining cost is easily made up for in the value of earlier release. Or along the route of higher quality frames, the same amount of power cost plus more in depth graphics is valuable to be seen as the pioneers in the field, plus having more visually appealing movies.
It is probable also very much what you're saying that hardware is a drop in the bucket compared to the amount they're making.
It's kinda like using an S3 Virge 8MB PCI video card and complaining about low Quake 3 framerates. Either turn down the textures and details and play the game at low res, upgrade, or play Doom. There's nothing wrong with Doom, just like there's nothing wrong with lightweight window managers.
I use several machines that cross the scale from Pentium Pro to Pentium 4 and Athlon XPs. My main desktop, a Pentium 4 with 1 GB of RAM runs Gnome 2 great, ran Gnome 1.4 great as well. Gnome 2 is actually faster than 1.4, and has been anywhere I've used it, regardless of memory size. At any rate, I want them to feel free to use resources as long as it doesn't bloat to the point where I have issues running a slew of programs at once. So far it hasn't and I'm a big fan of the improvements that keep going into Gnome. I also know better than to try putting Gnome on the Pentium Pro. It just isn't a great idea.
Beyond that, even most Pentium 2 systems have provisions for 256-512 MB of memory. The Pentium 2 400 I have at work ran Gnome 2 fine on 128 MB under conservative usage, ie not having 2 dozen programs open at once. It's kinda common sense that the machine will swap at that point. I ended up with more RAM, but that was because I was running RAM happy experiments. Gnome wasn't that big of a factor.
And RAM isn't exactly expensive either. I understand not wanting to pay any more on hardware to run programs, but again, if you want lightweight there are many, many other programs out there that are. They don't give you the moon on a stick, but are very reasonable alternatives for boxen that have small quantities of RAM.
When I saw it sitting there in my binder though, the first thought I had was, "Gee, don't see many encyclopedias around these days..."
I find strangers that just walk up to me on the street and start talking to be extremely creepy. Yes, I'm an introvert, but I have plenty of friends and hold conversations and even arguements with people I know without issue. What in the world posseses somebody to just strike up a conversation with me I'll never know.
I wear headphones when I'm out walking alone very often because I love listening to music and am not planning to interact with anybody. That doesn't mean I won't take them off if I see interaction coming up. I did so just last week to help somebody take a picture of themself against the Minneapolis skyline. No biggie, and it wasn't somebody randomly wanting to hear my life story or tell me theirs. The only time it has ever stopped interaction that I knew of and I didn't take them off is the religious nuts that stop everybody they think will listen. But I've stopped and played their games and debated philosophy with them too. At least I know why they're talking to random people.
I smile at people as I walk past them, and often they smile back. It's polite and unobtrusive. I'll never just randomly start a conversation though. People that do just weird me out way too much.
A subtitling is under no such pressure. If the subtitle takes its sweet time to say the same thing, or puts the sentence into the five words it really means, well, it does. You read, you comprehend, you get sound with the mouth moving, even if it isn't perfect. (You don't think the Japanese always lines up perfect, do you?)
There's also another thing that fansubbing does at least. That is to use a few Japanese words (well, their romanizations) in the sub itself. The average watcher is generally pretty clueless to it, so I don't think it'd be a mainstream (read: Disney) thing, but it adds a lot when these little things are carried over because they have a different feel to them than the English translations tend to. Especially the suffixes used with names, sensei, sama, chan, kun, etc.
When these are translated to what they mean, like little Yahiko in the Kenshin DVD subs (can't stand the English voices) instead of Yahiko-chan like in the fansubs it makes my skin crawl. Little Yahiko is not something that anybody'd think to say in English except maybe in the instance of "poor little Yahiko" which would have waaay been better in the Kenshin instances because chan is used as mocking Yahiko. Anywhere else and you don't really think to call a child little, so "Thank you little Chihiro" is right out, where "Arigato Chihiro-chan" is, well, proper (even if I may have killed the spelling). So it invariably gets dropped in the dubbing process. This leads to a different problem, you lose the information of the relation of two characters by not using these methods of address. Sama and sensei are terms of respect, and to call Yahiko by the term Yahiko-sama means you have respect for them. Instead it's just lost, or translated into a Mr. Yahiko, which in a lot of cases sounds dumb and doesn't imply the same respect.
Just my two cents on the whole dubbing process. And before anybody thinks I'm being too harsh, let me say I have a lot of respect for the people that dub in both the original Japanese and the English releases. I've been working on a fandub with some friends (bad idea, but fun to try) and, while it's coming along, it is incredibly time consuming and difficult to get the voices to sound consistantly in character, let alone capture what's being said in approximately the proper length of time and sound like you mean it. I imagine there are similar problems getting from Japanese script to English subs, just to make the lines sound right when reading, but I think there's less pressure to change things to fit the length of time that mouths are moving and thus the sub can be more true to the original Japanese.
Guess I'll get around to that letter crying to Lucasarts too. I was looking forward to this game. However, if anybody hasn't read a Sam and Max comic book I suggest picking one up. It might ease the pain a little bit.
Damn, and I'm in college still. I should've been able to guess that one. Guess grad school is out of the realm where many people feel the need to use 30 point wide fixed width fonts to demonstrate their incompetence. Plenty of ways to do that with content.
Just what I've heard, but it'd make a sort of logical sense as to why to stay away from Word if you're in law. Of course, I'm not sure why there'd be a hard word maximum and not just a page maximum. But there's a lot of quirks in law I'm not so sure of.
Nah, I got it and tried to make the reply as funny as possible while still addressing a real issue. Guess I didn't do too well there. :)
As for why most young people are involved in homicides, I remember doing the research for some Fark topic and finding that there's way more young people involved in killing or being killed than you'd expect, but most of it is gang related. Usually these crimes are perpetuated in the neighborhoods that are least likely to have a bunch of pasty white suburban children (excuse my stereotyping) imitating GTA. They've been living out the real thing without ever playing for a while.
When the pasty white suburban kid does go nuts and kill though it seems almost certain to be because the parents could not keep tabs on their child through the developmental stages. This is probably because the kids got themselves confused about real and fake when they were impressionable, completely at the fault of the parents. The same is really true that if parents could exercise a little more control even in bad neighborhoods that there would likely be less violence. I'm just going to be realistic about it and say that the gang neighborhoods are in a state where it's way too easy to propogate violent tendencies and the best move a parent could make is to leave if they can.
Some people may be sociopaths, but I doubt that all that many really are. It doesn't seem like a trait that would be prone to stick with society too well and, well, frankly isn't going to be expressed or not expressed because of a video game.
Sure, a person could be either amoral or a moron because they're influenced to kill by a video game. What made that person that way? I'd say it's likely that their parents made a few too many bad choices, didn't pay enough attention, or thought that children raise themselves.
It's high time victims in this country start suing parents of underage perpetrators of violent crime. Somebody's 16 year old kid tries to mow you down with their car? You should be able to sue the hell out of their parents for negligence of some sort. That'd teach parents to stop shifting blame for their being crappy moral compasses. Maybe then they'd start to pay attention to what is raising their children when they aren't. Be it video games, movies, beer commercials, or even the 10 o'clock news.
Yeah, I think that on an impressionable mind, especially that of somebody under 8 years of age, any sort of violent media could easily be misinterpreted. That's why there's parents. Adult entertainment is labeled as adult entertainment because it's for adults. If an adult cannot handle it then it's probably because of something they were taught as a child.
No, I think he means that sane people will have more fun playing the game than they might expect out of a sequel which probably will have only incremental improvements to its engine.
Expectations about stupid, homicidal people being stupid and homicidal are already high, regardless of their playing a video game or desire to have a scapegoat for their stupid and homicidal actions. As such, I'm pretty sure he was mentioning just the game aspect.
Beyond that, I think it provides an excellent answer to the "What can you bring to the company?" question. Being open they can inspect what the software you've codeveloped is like and see if they like it. It's win-win because you get to show off and the company gets to see the quality of the developer they're about to hire. Try to get that sort of perk from your last proprietary company.
Finally, open source projects provide something in a poor student's free time that they probably aren't apt to get during their college career. That's experience. I wrote a three hundred line program on my own for a class doesn't sound nearly as good as being responsible for even a few hundred lines in a much larger project, even if nobody really knows about either. And again, if it's a popular project then so much the better. What's the chances of working on a big project with no experience during college?
And these are just the personal benefits of having written some of the code. While working on the Linux kernel may not pay the rent, paying hundreds of dollars every time you want (or even need if you're a developer) to get a new version of some particular piece of software doesn't help pay it either. Think for a moment how much of a person's salary could go towards paying for the professional version of a proprietary operating system, office suite, database, and development studio. $2,500 down and about half of that every time there's an update, which seems to be every 2-3 years. Of course, last I checked that doesn't provide anything that would allow you to run your own webserver to try developing web applications, nor does it provide you with a source code repository for version tracking, nor could you run an email server if you wanted to try your hand at administering one of those. List keeps going.
There's a lot of benefit to everybody when there is free software at hand. There's plenty of software left to be written and companies will pay for that. But there's also a lot of software that already exists and isn't going to make anybody except the big companies money when it's sold while the company pays its workers to make more software. It is also likely cost people interested in the technical side of it more, since you can't just learn about it for free.
I actually like having a little stress in my life now and again. Things could be awfully dull without it. And the best part is triumphing over those stressful challenges and feeling good about it.
I couldn't imagine being stressed to the point of heart attack though. Maybe I manage it well, but it just needs to be set aside if it gets that bad. Failure isn't the worst thing in this case.
Apart from that, everybody keeps looking at stress as though it's a bad thing. I know some of my best work is done under the heavy pressure of stress. While a lot of stress certainly wouldn't be a healthy level for me to maintain, a bit of stress, even really intense stress, can be good for you and keep you from being complacent. I'd hate to lead a completely stress free life.
"Well, most people still have no use for these instructions, and we could have developed a set of 64 bit extensions to our own architecture on a whim, but in the spirit of preventing confusion about which extensions are in use on our x86 instruction set and to allow for easier forward progress when people do need 64 bit computing, we'll just implement what AMD threw together. You know, just because it's there, not because it's difficult. Why waste money on developing diverging 'standards'?"
That'd be my spin anyway, regardless of how wrong it may be.
While I think the Opteron is a great choice, and kudos to your company for getting the best of what they can right now, many companies are also either blindly loyal to Intel or don't want to adopt anything too "new" too quickly. The fact that a few linux distros and now some beta software from MS will take full advantage of the Opteron will probably have an effect too, despite the Opteron running 32-bit code very nicely. It's the uninformed managers that tend to make those decisions.
Again, congrats to your company which decided to jump on already. I'm still of the opinion that there will be plenty of 64-bit sales for Intel and this is no more than a minor scrape, completely detached from a favor for either company.
Also, it's been stated elsewhere that the Prescott cores already have the extensions, they just aren't enabled. So there's apparently no problem getting the extensions working on the chips, in the worst case they don't have to develop any of the extensions, and AMD has blown a bunch of cash raising interest in 64 bit processors.
I think Intel probably isn't going to be hurting so bad here.
It seemed for a while though, that Sun was at least trying to make Solaris a desktop friendly OS. They've been putting work into their office suite and have adopted Gnome as a WM in addition to CDE. Though from what I hear now, they're trying to make their Java Desktop which runs Linux their main desktop OS. Regardless, Sun is trying to take desktop share away from MS and is still around. Which is more than we can say about Be.
I won't claim that it has droves of supporters and users, but Solaris does run on x86 hardware and can (technically) be used as a desktop.