Free as in beer: True, but not much of a point if you have a job that pays at least minimum wage. I think I'd rather install Windows and have it work than try to figure out which of hundreds of distros and versions to use and getting one of them to actually work right on my system.
Eh? Well, at minimum wage, coming up with $300 (one week's pay before taxes) to buy a new copy of Win7 will be quite a struggle, but I get that you are exaggerating for effect. Anyway, there are really only 3 major players in the desktop Linux space: Fedora, Mint, and Ubuntu. Although since you say you are running Ubuntu, I think you already know that. I haven't used Fedora for many years because every time I do there are serious bugs and stability issues. I just don't have the patience or time to deal with it, so I've been using Ubuntu and have more or less been happy.
I don't use it because it's free. I'm past needing free as in beer software now. I use plenty of commercial software on my Linux box. I use it because it works well for me. As much as OS X is nice and flashy, and Windows has improved drastically over the years, they both still constantly frustrate me. The free as in speech/open, I do care about, but only in so much as it affects the seamless functioning of the system. I want open source drivers because they can be integrated well to work with the kernel, and kernel devs can patch and update them. I want open source libraries so that useful utilities that I use all the time can be written by college students without worrying about licensing. I want open codecs so that I don't have to worry about video and sound playback on my machine, and I don't have to use crappy commercial dvd software to watch movies. I want applications that use open formats to handle my data so that I don't have to worry about proprietary formats holding it hostage.
And finally, I don't want to be on an endless upgrade churn constantly buying new software versions for incremental improvements that still keep crappy legacy code around because the companies behind them don't want to invest the time and money to properly rewrite them to support the newer toolkits and technologies. Endnote X5 still uses the widget toolkit from Windows 2000 and doesn't properly support installing the toolbar for multiple users. And to get support for newer versions of Word and Windows, yep, gotta buy a new version. Filemaker Pro occasionally has bugs that don't mix well with OS X updates, mostly because they are still using unsupported and deprecated functions. When a problem comes up, do they issue a patch? Nope, gotta buy the new version. Photoshop only just recently, after many years and many versions on OS X, finally removed their last Carbon dependencies in the latest version. It's a >$1000 program! Why can't they properly update sooner? I really hate this about a lot (not all) of commercial software, and it doesn't seem to be a problem with open source software.
Stable and efficient: I'll believe that when somebody tells me why no kernel later than -33 will boot my system
No idea. But a very unusual problem. What kind of hardware setup and what version of Ubuntu do you have?
I just recently did a fresh Win7 install on a laptop. I never tried to install Vista, but XP was always a pain, with an installer from the mid-90s that never could recognize new hardware very well. Win7 in that respect, is very much improved. It was a breeze, fast and easy, the way Linux distributions have been for many years. However, after installing, it took several hours to get everything completely up to date. Windows Update is still, after >10yrs, the biggest piece of crap in existence. Linux, by contrast, takes about 15-30 minutes (depending on Internet and computer speed) to get all relevant updates (for everything, OS and software), with at most one reboot. That is something that Windows really needs. I hate installing Windows for this reason alone.
That completely unsubstantiated statement seems to come up in every discussion about Unity/Gnome3. Have you ever used a tablet? Unity/Gnome3 would be horrible interfaces for one. They are not designed to be for tablets. And yes, they are not "traditional" either, whatever that means.
Well, I did say there are good and bad IT departments at universities. I have only experienced good ones, though, and have never seen anything like you describe. There are network restrictions, yes (see my point #1), but a good IT department knows how to secure a network with essentially untrusted clients. So they don't need to excessively control the machines that are connected to the network, which is what the majority of the people were complaining about in the comments I read.
Also, I have used a lot of scientific software both for data processing and for running instrumentation, and have never had a problem with remote monitoring software interference. Occasionally anti-virus gets in the way, but the problems are usually minor, and IT would let us disable it if we really needed to. What was the problem in your case, I'm curious?
You know, it strikes me after reading many of these comments, that the IT structures most people seem to hate are in corporations. I wonder why more departments don't operate the way they do in universities. I mean, most universities have very large networks with independent and disconnected people all trying to do their own things. They all have very different IT needs, and the basic needs of the infrastructure are still there (stability and security). At the same time, the IT departments aren't heavily funded, so they have to make do with what they have. In these situations, I have found the IT departments (the good ones at least) manage fairly well. It basically amounts to a few key strategies:
1) The network (uptime and security) is the responsibility of ITS. So no rogue wireless access points, no dhcp servers, everybody has a controlled network account with a strict password policy, etc. In return, the users get a single stable network (wired and wireless) from which to do their work. They can get static ip addresses if they need them, domain names for their servers, firewall exceptions, vpn access, domain authentications, single sign-on, mailboxes, network storage, personal webpages, etc. If they need a new network drop they can have one installed, or if a port isn't working they can expect a network guy to take care of it. For the most part, it's an arrangement that works pretty well and I have seen little dispute over it.
2) Offices with specific software requirements and no time or desire to manage it themselves have IT-managed computers. The software people need is there. The computers work. No administrative access is given. No flexibility in software choice is given. If there is a problem, the IT guys respond quickly and efficiently.
3) Computer labs and classrooms are run differently based on the needs, but one of the more useful setups I have seen is where complete access to the computer is given for a session, but the the computer wipes and resets itself after a period of inactivity.
4) Individual users and departments are free to setup their computers however they wish. a) They can go the entirely independent route (most students/staff pick this one). Reasonable assistance from the IT guys can be expected, but it is understood that they are unable to help with everything and that you are on your own if you go against their recommendations. Any computer that connects to the network must conform to the network policy. Anti-virus/anti-malware/strong passwords aren't strictly required, but if the network scanner picks up suspicious activity your computer will be banned until it is fixed. If your network account gets compromised, it will also be shut off until the problem is fixed. b) They can go the semi-IT-managed route (many faculty pick this one), where IT sets up the computer for them based on their software and platform needs. They monitor backups and critical updates for you, and keep an administrative account on your machine to do this, but they don't restrict you from having administrative access to your own machine. If you screw up your machine, you understand that it is your time that is being lost and that, while IT will help you get it back up and running, they aren't able to drop everything else that they are doing and you may have to wait. This usually causes people to be a little more conservative with what they do on their computers. c) Or they can go the fully-IT-managed route (most general purpose workstations and equipment computers are configured this way). They typically have domain logons, no administrator access, and a strict software set. Additional software can be installed as needed, but it has to be done by IT. The primary requirement is that the systems be available for use and not suffer frequent unnecessary downtime.
It is not a homogenous one-size-fits-all setup for everyone, because it is understood that everybody has different needs. There is a balance between what a budget-strapped IT department can provide and what users need from the network
After reading this thread, I can only conclude a few things:
1) A large sample of Slashdotters claim to (but obviously don't) know anything about MBAs, what they do, or what they study. There even seems to be some confusion about the degree program itself. MBA (that is Masters in Business Administration) != anything vaguely business related (finance, economics, investment/stock mangement).
2) There seems to be a certain naiveté being expressed that a person holding a degree should be expressly proficient (thereby validating the existence of the degree). Every Slashdotter here knows (or should know), though, that isn't the case. There are plenty of horrible programmers (ahem "software engineers") that hold degrees from reputable institutions. But we don't disparage the whole field and its practitioners due to a few (or large number) of bad apples now, do we?
3) A point being made repeatably, albeit confusedly, is that running a successful business requires some luck, some opportunity, and a large number of skills (likely contributed by a large number of people). Those skilled people become good at what they do from a mixture of study (formal or otherwise) and experience (successful or otherwise). So a person who holds an MBA is not necessarily going to be a good businessman. Getting an MBA may provide him with some tools that will contribute to his success, but it doesn't guarantee it and it doesn't mean that you must have it to be successful.
The short of it is, there are good businessman and bad. Some of them have MBAs and some don't. Having an MBA does not mean you will be either, but many good businessman will find it useful to have an MBA as part of their formal training.
(footnote) There is also the issue that it seems to be assumed that an executive should necessarily have the interests of the company as his priority and not simply his personal gain. The MBA is often blamed for evidence of the latter, but there is no clear obvious link between MBA and greedy, soulless bastard (despite what many Slashdotters want to believe). You can really only blame our culture for that. And the degeneration of the stock market from a means to enable the growth of healthy companies into a mere gambling enterprise.
Maybe. Or maybe we just prefer an honest discussion. From my point-of-view, I don't own a Mac, but I really like their hardware. I think their products are well-built and have a very high quality that fully justifies their higher price. So I will recommend macbooks to people looking to buy a laptop. Now, somebody comes along and says macbooks are just overpriced and anything else is just as good, that "I'm brainwashed" for thinking they are higher quality, or that I only prefer macbooks because I want to be cool and hip, I take offense to that and will defend my not-brainwashed opinion. You don't see people claiming it is brainwashed to like Thinkpads, although there quite a few people who will stand by those. There are just a lot of people out there who are not content to simply dislike Apple and not buy their products, they have to also denigrate those who do. So the defensive reactions of those who buy, recommend, and are happy with Apple products is not entirely unsurprising to me.
The only Macbook that costs $2500 is the 17-inch Pro. Toshiba only sells a handful of 17-inch laptops, and based on what you say of the price, I'm guessing you have a Qosmio X775. I just randomly picked one, so let's go through it, shall we?
Case and form factor (or "build quality" as others are referring to) Weight: 7.5 lbs for the Toshiba vs. 6.6 lbs. for the Apple Thickness: not in the Toshiba specs, but from the pictures looks like about 2 inches vs. 1 inch for the Apple Battery life: 5 hours on the Toshiba (testing conditions not disclosed) vs. 7 hours for the Apple (50% display brightness browsing the web with the wireless card) Case material: plastic for the Toshiba vs. aluminum for the Apple Display material: plastic/polymer for the Toshiba vs. glass for the Apple
Display and Graphics Video adapter: NVidia GTX 1.5 Gb GDDR5 for the Toshiba vs. TWO graphics chipsets (AMD Radeon 1 Gb DDR5 + Intel HD3000, which it automatically switches between depending on workload, presumably to help the battery life) for the Apple Native resolution: 1600x900 for the Toshiba vs. 1920x1200 for the Apple Dual display: no mention of this on the Toshiba, but maximum external display resolution for the Apple is 2560x1600, which it can support while also driving the native resolution of the built-in display Display: nothing about this for the Toshiba, but Apple is well-known for their high-quality displays --from notebookcheck.net there are a few more specs (similar, but not the exact models)
avg. brightness: 192 for the Toshiba vs. 290 for the Apple
contrast: 250:1 for the Toshiba vs. 578:1 for the Apple
illumination: 77% for the Toshiba vs. 70% for the Apple
blacks: 0.9 cd/m^2 for the Toshiba vs. 0.6 cd/m^2 for the Apple
Memory, cpu, hard disk RAM: 6 Gb DDR3 for the Toshiba vs. 4 Gb DDR3 for the Apple Processor: Intel Core i5 2.4 GHz for the Toshiba vs. Intel quad-core i7 2.4Ghz for the Apple Hard disk: 640 Gb 7200 rpm for the Toshiba vs. 750 Gb 5400 rpm (optional 750 Gb 7200 rpm for $50) for the Apple
Everything else LAN: gigabit for both WLAN: b/g/n for the Toshiba vs. a/b/g/n for the Apple Bluetooth: 3.0 for the Toshiba vs. 2.1 for the Apple Smart card reader: none for the Toshiba vs. SDXC for the Apple Ports: USB 2/3 for the Toshiba vs. USB 2 + Firewire for the Apple External display port: HDMI for the Toshiba vs. Thunderbolt for the Apple, which supports VGA/DVI/dual-link DVI/HDMI
This is just a careful spec-to-spec comparison of the two. No need to pull in anecdotes or "feelings about quality". The Apple clearly has more higher-spec'd hardware for that extra $1400. You can certainly argue that the Toshiba is "good enough". That's fine, a personal decision. But the Apple is more expensive because it has more, and the quality of some of the components is higher, plain and simple.
Even in the ads that focus on people using Apple products in various way, it's always cool people using them.
Mmmm...I don't think it is any more exaggerated than other technology ads. Microsoft seems to like adds with people dancing around carrying colorful signs. I guess because using Windows/Office is supposed to make you feel cheerful and happy. Other ads similarly appeal to a desired image (think about the the stay at home moms who are just thrilled to be mopping the floor with Mr. Clean's Super Mop Formula for some reason). So while it is true that Apple wants to portray happiness or style or hipness in association with their products, I don't think it is any different from anybody else, and attributing "brainwashing" to Apple's success is just overstatement.
Of course at the big Uni's the Uni took 50% to 60% off the top to cover operational expenses, so every grant application had to include a justification for double the amount of money actually needed (since the grants rarely paid for operational expenses), hidden in the cost structure.
Actually, I'm pretty sure the major granting agencies know that the institution is going to take 50% and just assume your actual research budget is half of what you are asking for. The NIH grants make it explicit, so no guessing needed. Also, I wouldn't really call it shenanigans. The funding climate is extremely conservative right now. You have to have a track record to get funding, and that means lots of preliminary data. And, since that requires actual experiments to be done, that means using a portion of existing funding to kickstart the next project. This is the principle behind startup packages at universities. A new faculty member has no chance of getting a grant until they publish a few papers, so they have to run their labs entirely on the startup funds they are given for the first several years.
Do you really think so? At 10-20 times the cost of a regular battery, your phone battery would be $400-$800, and you want to buy two of them? And how long do you keep a phone for anyway? Certainly not for life....
The technology is good for some things, for sure. I'm thinking hybrids/EVs and power plant stations.
It depends, a lot, on where you are. If you are in the mountains of New Hampshire or Vermont, then you are probably right. And it's ok. If there really aren't very many people, a few cars on the road isn't a big deal. However, I live in College Station now, a college town. A high speed rail line that could get to Houston or Austin in 20 minutes, or to San Antonio or Dallas in 40 minutes, would be extremely popular. There is a decent bus system to get people to the station. And, since it is Texas, there is no shortage of land to build a gigantic parking lot for the people who aren't connected to the bus system or who want to drive anyway. In the destination cities, there are plenty of buses, taxis, and relatives, so a car wouldn't be absolutely necessary. In some cases it would, and people would drive, but I think a lot of people would really like the train, especially for the airport traffic. And once you have that, connecting to other nearby cities isn't such a large stretch either.
Dude, chill out. We are having a discussion here. No need to be so passionate.
Stop reading more into what I am saying than what I am saying. I already said a couple of times that a $250k FTE is in the range for a tenured professor (or mid-career scientist using your terminology). The point behind the postdoc diversion is that it isn't the tenured professors doing the lab work. When you need to hire someone to work on a project in your lab, you don't hire the equivalent of a tenured professor, you hire a postdoc or a grad student. When you apply for a grant, the cost you budget for personnel is usually taking into consideration the cost of postdocs and grad students, not "staff scientists." Heck, the salary of a tenured professor is usually in the employment contract with the university and has nothing to with any grants they might apply for. It is true, however, that some labs with secure funding will budget for a couple of "permadocs" at higher salaries, I don't dispute that.
As for what those salaries actually are, I only have my own experience as a reference. There is no database of postdoc salaries that I am aware of. What I know is that the NIH standard is a point of reference. The NIH, of course, pays at those rates and if you are lucky enough to get a funded fellowship you can expect to be paid at those rates. But those rates are too costly for many labs and so they pay less. And if you aren't in one of the big NIH fields (say, Ecology), you can expect to be paid a lot less. If physicists and engineers can regularly secure $45k+/yr in postdoc salaries, that's great for them, but it's a rarity in the life sciences.
The point behind all of this is that the money being offered is useful. It is not ideal, but it is useful, and not an atypical amount to apply for in a supportive grant.
Read my whole post. That is what NIH recommends. Not what labs actually pay. Even if the FTE is closer to $60k, which is what I conceded earlier, that is not even close to the $250k you are standing by. Even if you take the highest NIH salary and use your doubling estimate for FTE, that is $100k, also not even close to $250k. I'm sorry, but your claim is just ridiculous. Like I said earlier, tenured faculty might make that much, but not the majority.
I'm not writing off anything less than $1M as chump change, and I'm not writing off $40k as chump change. But I'm not kidding myself about how much it costs just to get people in the lab, and I think you are underestimating it significantly.
Getting people into the lab is expensive, I'm not disputing that. But there is a lot you can do with $40k that doesn't involve hiring personnel.
Your Caltech average salary is sourcing 10 postdocs at Caltech. There are a lot more than that, I am sure, so I think that is a fairly selective sampling. Here are the NIH guidelines for postdoc salaries in 2010,
Most labs pay less than the NIH, unless they have a lot of funding. Also, postdocs are usually temporary positions lasting 1-3 yrs. So the top of the pay scale is often not reached. Now, I am using life science salaries as my benchmark because that is all I have experience with, but it is quite possible that other fields pay more.
I understand your point: science is expensive. But writing off anything less than $1 million as chump change is not very constructive. First of all, even if you can only hire one extra person, that is one extra person for one year. Other labs might use it to upgrade some equipment or something. Every little bit helps. $40k, for example, will buy a new rotor for a centrifuge, or a shaking incubator, or an hplc...all great things that don't usually fit into the day-to-day budgets of most labs. Second, nobody is suggesting that this funding will be enough to completely sustain a lab. Most labs need several grants going at any given time to keep the lights on. This is just a boost, and you can do a lot of experiments with $40k.
$40k is a lot for a postdoc, and FTE isn't double. Maybe $60k, still a lot less than the $250k you were saying earlier. And yes, grad students are expensive.
Anyway, postdocs and grad students DO the research. Yes, the PI "supervises", but you will have one PI for a group of 10 or more postdocs/grad students. No, it's not ideal, but that is just how it works.
Your original point, I think, was that personnel are the most expensive part of doing research. That is definitely true. However, if the total funding is $250k as the grandparent-grandparent was saying, that is comparable to 1 yr. of RO1 funding from NIH.
Hahahaaaahhahahahahaaaa. Wow, what a great joke. This FTE is for who, exactly? A tenured PI in a place with a high COLA, maybe. A typical FTE for a grad student/postdoc is around 40k/yr, in the life sciences, which is usually the highest paid.
Yes, I agree, you are just reinforcing my point. Linux--or Mac, this isn't just about Linux--is not a suitable choice universally. Some actually NEED RAdmin, for example. Other do not, though, and if they don't one of the Linux alternatives would suit them just fine. The choice then lies between: do I really need RAdmin, or do I prefer to use Linux if it is possible. Like I said earlier, this depends on how you value cross-platform software over other things. If RAdmin has a killer feature you really need then you rightly prefer it, and cannot switch for that reason. Otherwise if you value certain features of Linux and other Linux-software, you might decide RAdmin isn't that important and you can switch to using an alternative.
What the hell? You do know mysql runs on windows, right? And that other databases like oracle run on linux? That is a pretty invalid point to be making.
I don't agree. Active directory is nice, for sure, but it is overkill for smallish networks. Central account management, sure, you can use ldap/Kerberos. But the strength of active directory is all of the remote client management and group policy stuff. Very useful in some situations, but certainly not critical in many.
There is no intrinsic reason to favor cross-platform applications. You have to value them, for one reason or another. I, for example, appreciate their flexibility and think they tend to be more robust than a lot of uniplatform software. The thread started out talking about adobe vs gimp. One person thought adobe was necessary (maybe it was), the other thought gimp was sufficient and much more wallet friendly. Both are pragmatic decisions, although some personal preference probably biases the decision somewhat. So if "the management" doesn't value cross-platform software, they won't use it. And that's fine, but it doesn't mean the option to use cross-platform software doesn't exist if they did value it.
Well, yeah, if you spec distinctly Windows-only software, you're not going to run it on Linux (or Mac for that matter). If you make some effort to be flexible, you can find a lot of great cross-platform software to suit your purposes. For example, Radmin like functionality can be achieved with vnc, freenx, and even Windows rdc. There are tons of videoconferencing apps, but Skype is probably the most well known. On Linux there is also the excellent ekiga. For silverlight/.net, you are pretty much screwed. But then if you really need windows software, you need to run windows. That has always been the case.
Keyword may that you dropped there. Most installers still set it up that way by default. I know I still prefer to have it that way even if it's not the default. In any case, if you know that there is a possibility that/usr and / will be on different filesystems (which there definitely is), you need to make sure your system will still boot if/usr can't be mounted (ie: init scripts can't depend on binaries or libraries in/usr).
No, I think you are confusing the compressed files/folders function (right-click, send to compressed folder) with the compression filesystem property (right-click, Properties, Advanced). The former is just a userspace program, like Winzip, that makes archives. The latter is exactly what you described in your second paragraph. The filesystem transparently compresses files in small clusters, and it suffers from fragmentation problems, like you also mention.
I was quoting somebody farther up in the thread...not you, I guess. Menus are easy to scan quickly, yes, unless you have a lot of entries. I can find Libreoffice Writer relatively quickly because it is, sensibly, placed under Applications|Office. The Terminal, however, seems to move around with every release because it doesn't inherently belong to any of the categories available. And if I don't already know where it is, looking for it is a random hunt through menus (is it under Programming, System Administration, System Tools,...). Don't you remember any of the usability studies Red Hat conducted a few years ago? People were trying to figure out how to check their email...it wasn't obvious that they had to open an application named "Evolution" that was placed under "Internet" instead of "Office" or "Messaging" or "Groupware", to cite just one example. They eventually figured it out, but it took them longer. I'm sure you've been in the situation where you have to go hunting through menus to find what you are looking for before (maybe you don't remember where it is or what it's called, or maybe you've never used the function before so you just have to randomly guess).
The rows of icons line by line isn't inherently faster, no. It's the particular implementation that makes it faster. Combined with the search, you only have to scan a handful (maybe five) icons, as opposed to running through an entire menu structure. The search learns from your selections too, so if you search for "Te" often and select "Terminal", it will start showing up at the start of the search results. If you just display all of the installed applications in rows then no, it isn't faster. Structure and organization is arbitrary. Your applications aren't "structured." They are all dumped into/usr/bin or/usr/X11/bin. Somebody came along and said, "let's help people try to find these easier," and they came up with the Applications menu, which is just a set of links to the actual applications. If you think about it, it is just a "search" that has been done by somebody else, and the results have been permanently stored in.desktop files.
The search box is just an extension of the idea. It let's you perform your own arbitrary search instead of being constrained to somebody else's. Search has to be done right, obviously. Can't speak to your Windows problem because I try to avoid it as much as possible. There are a lot of good search implementations, though...GMail and Spotlight are two that come to mind immediately. I see search as supplemental to organization, not a replacement for it (although some will use it that way). I can painstakingly organize all of my files into carefully labeled folders. But search will still help me find things faster than hunting through all of my folders looking for something I last worked on a year ago.
Slashdot Mirror
Free as in beer: True, but not much of a point if you have a job that pays at least minimum wage. I think I'd rather install Windows and have it work than try to figure out which of hundreds of distros and versions to use and getting one of them to actually work right on my system.
Eh? Well, at minimum wage, coming up with $300 (one week's pay before taxes) to buy a new copy of Win7 will be quite a struggle, but I get that you are exaggerating for effect. Anyway, there are really only 3 major players in the desktop Linux space: Fedora, Mint, and Ubuntu. Although since you say you are running Ubuntu, I think you already know that. I haven't used Fedora for many years because every time I do there are serious bugs and stability issues. I just don't have the patience or time to deal with it, so I've been using Ubuntu and have more or less been happy.
I don't use it because it's free. I'm past needing free as in beer software now. I use plenty of commercial software on my Linux box. I use it because it works well for me. As much as OS X is nice and flashy, and Windows has improved drastically over the years, they both still constantly frustrate me. The free as in speech/open, I do care about, but only in so much as it affects the seamless functioning of the system. I want open source drivers because they can be integrated well to work with the kernel, and kernel devs can patch and update them. I want open source libraries so that useful utilities that I use all the time can be written by college students without worrying about licensing. I want open codecs so that I don't have to worry about video and sound playback on my machine, and I don't have to use crappy commercial dvd software to watch movies. I want applications that use open formats to handle my data so that I don't have to worry about proprietary formats holding it hostage.
And finally, I don't want to be on an endless upgrade churn constantly buying new software versions for incremental improvements that still keep crappy legacy code around because the companies behind them don't want to invest the time and money to properly rewrite them to support the newer toolkits and technologies. Endnote X5 still uses the widget toolkit from Windows 2000 and doesn't properly support installing the toolbar for multiple users. And to get support for newer versions of Word and Windows, yep, gotta buy a new version. Filemaker Pro occasionally has bugs that don't mix well with OS X updates, mostly because they are still using unsupported and deprecated functions. When a problem comes up, do they issue a patch? Nope, gotta buy the new version. Photoshop only just recently, after many years and many versions on OS X, finally removed their last Carbon dependencies in the latest version. It's a >$1000 program! Why can't they properly update sooner? I really hate this about a lot (not all) of commercial software, and it doesn't seem to be a problem with open source software.
Stable and efficient: I'll believe that when somebody tells me why no kernel later than -33 will boot my system
No idea. But a very unusual problem. What kind of hardware setup and what version of Ubuntu do you have?
I just recently did a fresh Win7 install on a laptop. I never tried to install Vista, but XP was always a pain, with an installer from the mid-90s that never could recognize new hardware very well. Win7 in that respect, is very much improved. It was a breeze, fast and easy, the way Linux distributions have been for many years. However, after installing, it took several hours to get everything completely up to date. Windows Update is still, after >10yrs, the biggest piece of crap in existence. Linux, by contrast, takes about 15-30 minutes (depending on Internet and computer speed) to get all relevant updates (for everything, OS and software), with at most one reboot. That is something that Windows really needs. I hate installing Windows for this reason alone.
That completely unsubstantiated statement seems to come up in every discussion about Unity/Gnome3. Have you ever used a tablet? Unity/Gnome3 would be horrible interfaces for one. They are not designed to be for tablets. And yes, they are not "traditional" either, whatever that means.
Well, I did say there are good and bad IT departments at universities. I have only experienced good ones, though, and have never seen anything like you describe. There are network restrictions, yes (see my point #1), but a good IT department knows how to secure a network with essentially untrusted clients. So they don't need to excessively control the machines that are connected to the network, which is what the majority of the people were complaining about in the comments I read.
Also, I have used a lot of scientific software both for data processing and for running instrumentation, and have never had a problem with remote monitoring software interference. Occasionally anti-virus gets in the way, but the problems are usually minor, and IT would let us disable it if we really needed to. What was the problem in your case, I'm curious?
You know, it strikes me after reading many of these comments, that the IT structures most people seem to hate are in corporations. I wonder why more departments don't operate the way they do in universities. I mean, most universities have very large networks with independent and disconnected people all trying to do their own things. They all have very different IT needs, and the basic needs of the infrastructure are still there (stability and security). At the same time, the IT departments aren't heavily funded, so they have to make do with what they have. In these situations, I have found the IT departments (the good ones at least) manage fairly well. It basically amounts to a few key strategies:
1) The network (uptime and security) is the responsibility of ITS. So no rogue wireless access points, no dhcp servers, everybody has a controlled network account with a strict password policy, etc. In return, the users get a single stable network (wired and wireless) from which to do their work. They can get static ip addresses if they need them, domain names for their servers, firewall exceptions, vpn access, domain authentications, single sign-on, mailboxes, network storage, personal webpages, etc. If they need a new network drop they can have one installed, or if a port isn't working they can expect a network guy to take care of it. For the most part, it's an arrangement that works pretty well and I have seen little dispute over it.
2) Offices with specific software requirements and no time or desire to manage it themselves have IT-managed computers. The software people need is there. The computers work. No administrative access is given. No flexibility in software choice is given. If there is a problem, the IT guys respond quickly and efficiently.
3) Computer labs and classrooms are run differently based on the needs, but one of the more useful setups I have seen is where complete access to the computer is given for a session, but the the computer wipes and resets itself after a period of inactivity.
4) Individual users and departments are free to setup their computers however they wish. a) They can go the entirely independent route (most students/staff pick this one). Reasonable assistance from the IT guys can be expected, but it is understood that they are unable to help with everything and that you are on your own if you go against their recommendations. Any computer that connects to the network must conform to the network policy. Anti-virus/anti-malware/strong passwords aren't strictly required, but if the network scanner picks up suspicious activity your computer will be banned until it is fixed. If your network account gets compromised, it will also be shut off until the problem is fixed. b) They can go the semi-IT-managed route (many faculty pick this one), where IT sets up the computer for them based on their software and platform needs. They monitor backups and critical updates for you, and keep an administrative account on your machine to do this, but they don't restrict you from having administrative access to your own machine. If you screw up your machine, you understand that it is your time that is being lost and that, while IT will help you get it back up and running, they aren't able to drop everything else that they are doing and you may have to wait. This usually causes people to be a little more conservative with what they do on their computers. c) Or they can go the fully-IT-managed route (most general purpose workstations and equipment computers are configured this way). They typically have domain logons, no administrator access, and a strict software set. Additional software can be installed as needed, but it has to be done by IT. The primary requirement is that the systems be available for use and not suffer frequent unnecessary downtime.
It is not a homogenous one-size-fits-all setup for everyone, because it is understood that everybody has different needs. There is a balance between what a budget-strapped IT department can provide and what users need from the network
After reading this thread, I can only conclude a few things:
1) A large sample of Slashdotters claim to (but obviously don't) know anything about MBAs, what they do, or what they study. There even seems to be some confusion about the degree program itself. MBA (that is Masters in Business Administration) != anything vaguely business related (finance, economics, investment/stock mangement).
2) There seems to be a certain naiveté being expressed that a person holding a degree should be expressly proficient (thereby validating the existence of the degree). Every Slashdotter here knows (or should know), though, that isn't the case. There are plenty of horrible programmers (ahem "software engineers") that hold degrees from reputable institutions. But we don't disparage the whole field and its practitioners due to a few (or large number) of bad apples now, do we?
3) A point being made repeatably, albeit confusedly, is that running a successful business requires some luck, some opportunity, and a large number of skills (likely contributed by a large number of people). Those skilled people become good at what they do from a mixture of study (formal or otherwise) and experience (successful or otherwise). So a person who holds an MBA is not necessarily going to be a good businessman. Getting an MBA may provide him with some tools that will contribute to his success, but it doesn't guarantee it and it doesn't mean that you must have it to be successful.
The short of it is, there are good businessman and bad. Some of them have MBAs and some don't. Having an MBA does not mean you will be either, but many good businessman will find it useful to have an MBA as part of their formal training.
(footnote) There is also the issue that it seems to be assumed that an executive should necessarily have the interests of the company as his priority and not simply his personal gain. The MBA is often blamed for evidence of the latter, but there is no clear obvious link between MBA and greedy, soulless bastard (despite what many Slashdotters want to believe). You can really only blame our culture for that. And the degeneration of the stock market from a means to enable the growth of healthy companies into a mere gambling enterprise.
Maybe. Or maybe we just prefer an honest discussion. From my point-of-view, I don't own a Mac, but I really like their hardware. I think their products are well-built and have a very high quality that fully justifies their higher price. So I will recommend macbooks to people looking to buy a laptop. Now, somebody comes along and says macbooks are just overpriced and anything else is just as good, that "I'm brainwashed" for thinking they are higher quality, or that I only prefer macbooks because I want to be cool and hip, I take offense to that and will defend my not-brainwashed opinion. You don't see people claiming it is brainwashed to like Thinkpads, although there quite a few people who will stand by those. There are just a lot of people out there who are not content to simply dislike Apple and not buy their products, they have to also denigrate those who do. So the defensive reactions of those who buy, recommend, and are happy with Apple products is not entirely unsurprising to me.
The only Macbook that costs $2500 is the 17-inch Pro. Toshiba only sells a handful of 17-inch laptops, and based on what you say of the price, I'm guessing you have a Qosmio X775. I just randomly picked one, so let's go through it, shall we?
Case and form factor (or "build quality" as others are referring to)
Weight: 7.5 lbs for the Toshiba vs. 6.6 lbs. for the Apple
Thickness: not in the Toshiba specs, but from the pictures looks like about 2 inches vs. 1 inch for the Apple
Battery life: 5 hours on the Toshiba (testing conditions not disclosed) vs. 7 hours for the Apple (50% display brightness browsing the web with the wireless card)
Case material: plastic for the Toshiba vs. aluminum for the Apple
Display material: plastic/polymer for the Toshiba vs. glass for the Apple
Display and Graphics
Video adapter: NVidia GTX 1.5 Gb GDDR5 for the Toshiba vs. TWO graphics chipsets (AMD Radeon 1 Gb DDR5 + Intel HD3000, which it automatically switches between depending on workload, presumably to help the battery life) for the Apple
Native resolution: 1600x900 for the Toshiba vs. 1920x1200 for the Apple
Dual display: no mention of this on the Toshiba, but maximum external display resolution for the Apple is 2560x1600, which it can support while also driving the native resolution of the built-in display
Display: nothing about this for the Toshiba, but Apple is well-known for their high-quality displays
--from notebookcheck.net there are a few more specs (similar, but not the exact models)
avg. brightness: 192 for the Toshiba vs. 290 for the Apple
contrast: 250:1 for the Toshiba vs. 578:1 for the Apple
illumination: 77% for the Toshiba vs. 70% for the Apple
blacks: 0.9 cd/m^2 for the Toshiba vs. 0.6 cd/m^2 for the Apple
Memory, cpu, hard disk
RAM: 6 Gb DDR3 for the Toshiba vs. 4 Gb DDR3 for the Apple
Processor: Intel Core i5 2.4 GHz for the Toshiba vs. Intel quad-core i7 2.4Ghz for the Apple
Hard disk: 640 Gb 7200 rpm for the Toshiba vs. 750 Gb 5400 rpm (optional 750 Gb 7200 rpm for $50) for the Apple
Everything else
LAN: gigabit for both
WLAN: b/g/n for the Toshiba vs. a/b/g/n for the Apple
Bluetooth: 3.0 for the Toshiba vs. 2.1 for the Apple
Smart card reader: none for the Toshiba vs. SDXC for the Apple
Ports: USB 2/3 for the Toshiba vs. USB 2 + Firewire for the Apple
External display port: HDMI for the Toshiba vs. Thunderbolt for the Apple, which supports VGA/DVI/dual-link DVI/HDMI
This is just a careful spec-to-spec comparison of the two. No need to pull in anecdotes or "feelings about quality". The Apple clearly has more higher-spec'd hardware for that extra $1400. You can certainly argue that the Toshiba is "good enough". That's fine, a personal decision. But the Apple is more expensive because it has more, and the quality of some of the components is higher, plain and simple.
Even in the ads that focus on people using Apple products in various way, it's always cool people using them.
Mmmm...I don't think it is any more exaggerated than other technology ads. Microsoft seems to like adds with people dancing around carrying colorful signs. I guess because using Windows/Office is supposed to make you feel cheerful and happy. Other ads similarly appeal to a desired image (think about the the stay at home moms who are just thrilled to be mopping the floor with Mr. Clean's Super Mop Formula for some reason). So while it is true that Apple wants to portray happiness or style or hipness in association with their products, I don't think it is any different from anybody else, and attributing "brainwashing" to Apple's success is just overstatement.
Of course at the big Uni's the Uni took 50% to 60% off the top to cover operational expenses, so every grant application had to include a justification for double the amount of money actually needed (since the grants rarely paid for operational expenses), hidden in the cost structure.
Actually, I'm pretty sure the major granting agencies know that the institution is going to take 50% and just assume your actual research budget is half of what you are asking for. The NIH grants make it explicit, so no guessing needed. Also, I wouldn't really call it shenanigans. The funding climate is extremely conservative right now. You have to have a track record to get funding, and that means lots of preliminary data. And, since that requires actual experiments to be done, that means using a portion of existing funding to kickstart the next project. This is the principle behind startup packages at universities. A new faculty member has no chance of getting a grant until they publish a few papers, so they have to run their labs entirely on the startup funds they are given for the first several years.
Do you really think so? At 10-20 times the cost of a regular battery, your phone battery would be $400-$800, and you want to buy two of them? And how long do you keep a phone for anyway? Certainly not for life....
The technology is good for some things, for sure. I'm thinking hybrids/EVs and power plant stations.
It depends, a lot, on where you are. If you are in the mountains of New Hampshire or Vermont, then you are probably right. And it's ok. If there really aren't very many people, a few cars on the road isn't a big deal. However, I live in College Station now, a college town. A high speed rail line that could get to Houston or Austin in 20 minutes, or to San Antonio or Dallas in 40 minutes, would be extremely popular. There is a decent bus system to get people to the station. And, since it is Texas, there is no shortage of land to build a gigantic parking lot for the people who aren't connected to the bus system or who want to drive anyway. In the destination cities, there are plenty of buses, taxis, and relatives, so a car wouldn't be absolutely necessary. In some cases it would, and people would drive, but I think a lot of people would really like the train, especially for the airport traffic. And once you have that, connecting to other nearby cities isn't such a large stretch either.
Dude, chill out. We are having a discussion here. No need to be so passionate.
Stop reading more into what I am saying than what I am saying. I already said a couple of times that a $250k FTE is in the range for a tenured professor (or mid-career scientist using your terminology). The point behind the postdoc diversion is that it isn't the tenured professors doing the lab work. When you need to hire someone to work on a project in your lab, you don't hire the equivalent of a tenured professor, you hire a postdoc or a grad student. When you apply for a grant, the cost you budget for personnel is usually taking into consideration the cost of postdocs and grad students, not "staff scientists." Heck, the salary of a tenured professor is usually in the employment contract with the university and has nothing to with any grants they might apply for. It is true, however, that some labs with secure funding will budget for a couple of "permadocs" at higher salaries, I don't dispute that.
As for what those salaries actually are, I only have my own experience as a reference. There is no database of postdoc salaries that I am aware of. What I know is that the NIH standard is a point of reference. The NIH, of course, pays at those rates and if you are lucky enough to get a funded fellowship you can expect to be paid at those rates. But those rates are too costly for many labs and so they pay less. And if you aren't in one of the big NIH fields (say, Ecology), you can expect to be paid a lot less. If physicists and engineers can regularly secure $45k+/yr in postdoc salaries, that's great for them, but it's a rarity in the life sciences.
The point behind all of this is that the money being offered is useful. It is not ideal, but it is useful, and not an atypical amount to apply for in a supportive grant.
Read my whole post. That is what NIH recommends. Not what labs actually pay. Even if the FTE is closer to $60k, which is what I conceded earlier, that is not even close to the $250k you are standing by. Even if you take the highest NIH salary and use your doubling estimate for FTE, that is $100k, also not even close to $250k. I'm sorry, but your claim is just ridiculous. Like I said earlier, tenured faculty might make that much, but not the majority.
I'm not writing off anything less than $1M as chump change, and I'm not writing off $40k as chump change. But I'm not kidding myself about how much it costs just to get people in the lab, and I think you are underestimating it significantly.
Getting people into the lab is expensive, I'm not disputing that. But there is a lot you can do with $40k that doesn't involve hiring personnel.
Your Caltech average salary is sourcing 10 postdocs at Caltech. There are a lot more than that, I am sure, so I think that is a fairly selective sampling. Here are the NIH guidelines for postdoc salaries in 2010,
http://grants.nih.gov/grants/guide/notice-files/not-od-10-047.html
Most labs pay less than the NIH, unless they have a lot of funding. Also, postdocs are usually temporary positions lasting 1-3 yrs. So the top of the pay scale is often not reached. Now, I am using life science salaries as my benchmark because that is all I have experience with, but it is quite possible that other fields pay more.
I understand your point: science is expensive. But writing off anything less than $1 million as chump change is not very constructive. First of all, even if you can only hire one extra person, that is one extra person for one year. Other labs might use it to upgrade some equipment or something. Every little bit helps. $40k, for example, will buy a new rotor for a centrifuge, or a shaking incubator, or an hplc...all great things that don't usually fit into the day-to-day budgets of most labs. Second, nobody is suggesting that this funding will be enough to completely sustain a lab. Most labs need several grants going at any given time to keep the lights on. This is just a boost, and you can do a lot of experiments with $40k.
$40k is a lot for a postdoc, and FTE isn't double. Maybe $60k, still a lot less than the $250k you were saying earlier. And yes, grad students are expensive.
Anyway, postdocs and grad students DO the research. Yes, the PI "supervises", but you will have one PI for a group of 10 or more postdocs/grad students. No, it's not ideal, but that is just how it works.
Your original point, I think, was that personnel are the most expensive part of doing research. That is definitely true. However, if the total funding is $250k as the grandparent-grandparent was saying, that is comparable to 1 yr. of RO1 funding from NIH.
Hahahaaaahhahahahahaaaa. Wow, what a great joke. This FTE is for who, exactly? A tenured PI in a place with a high COLA, maybe. A typical FTE for a grad student /postdoc is around 40k/yr, in the life sciences, which is usually the highest paid.
Yes, I agree, you are just reinforcing my point. Linux--or Mac, this isn't just about Linux--is not a suitable choice universally. Some actually NEED RAdmin, for example. Other do not, though, and if they don't one of the Linux alternatives would suit them just fine. The choice then lies between: do I really need RAdmin, or do I prefer to use Linux if it is possible. Like I said earlier, this depends on how you value cross-platform software over other things. If RAdmin has a killer feature you really need then you rightly prefer it, and cannot switch for that reason. Otherwise if you value certain features of Linux and other Linux-software, you might decide RAdmin isn't that important and you can switch to using an alternative.
What the hell? You do know mysql runs on windows, right? And that other databases like oracle run on linux? That is a pretty invalid point to be making.
I don't agree. Active directory is nice, for sure, but it is overkill for smallish networks. Central account management, sure, you can use ldap/Kerberos. But the strength of active directory is all of the remote client management and group policy stuff. Very useful in some situations, but certainly not critical in many.
There is no intrinsic reason to favor cross-platform applications. You have to value them, for one reason or another. I, for example, appreciate their flexibility and think they tend to be more robust than a lot of uniplatform software. The thread started out talking about adobe vs gimp. One person thought adobe was necessary (maybe it was), the other thought gimp was sufficient and much more wallet friendly. Both are pragmatic decisions, although some personal preference probably biases the decision somewhat. So if "the management" doesn't value cross-platform software, they won't use it. And that's fine, but it doesn't mean the option to use cross-platform software doesn't exist if they did value it.
Well, yeah, if you spec distinctly Windows-only software, you're not going to run it on Linux (or Mac for that matter). If you make some effort to be flexible, you can find a lot of great cross-platform software to suit your purposes. For example, Radmin like functionality can be achieved with vnc, freenx, and even Windows rdc. There are tons of videoconferencing apps, but Skype is probably the most well known. On Linux there is also the excellent ekiga. For silverlight/.net, you are pretty much screwed. But then if you really need windows software, you need to run windows. That has always been the case.
Keyword may that you dropped there. Most installers still set it up that way by default. I know I still prefer to have it that way even if it's not the default. In any case, if you know that there is a possibility that /usr and / will be on different filesystems (which there definitely is), you need to make sure your system will still boot if /usr can't be mounted (ie: init scripts can't depend on binaries or libraries in /usr).
No, I think you are confusing the compressed files/folders function (right-click, send to compressed folder) with the compression filesystem property (right-click, Properties, Advanced). The former is just a userspace program, like Winzip, that makes archives. The latter is exactly what you described in your second paragraph. The filesystem transparently compresses files in small clusters, and it suffers from fragmentation problems, like you also mention.
I was quoting somebody farther up in the thread...not you, I guess. Menus are easy to scan quickly, yes, unless you have a lot of entries. I can find Libreoffice Writer relatively quickly because it is, sensibly, placed under Applications|Office. The Terminal, however, seems to move around with every release because it doesn't inherently belong to any of the categories available. And if I don't already know where it is, looking for it is a random hunt through menus (is it under Programming, System Administration, System Tools, ...). Don't you remember any of the usability studies Red Hat conducted a few years ago? People were trying to figure out how to check their email...it wasn't obvious that they had to open an application named "Evolution" that was placed under "Internet" instead of "Office" or "Messaging" or "Groupware", to cite just one example. They eventually figured it out, but it took them longer. I'm sure you've been in the situation where you have to go hunting through menus to find what you are looking for before (maybe you don't remember where it is or what it's called, or maybe you've never used the function before so you just have to randomly guess).
The rows of icons line by line isn't inherently faster, no. It's the particular implementation that makes it faster. Combined with the search, you only have to scan a handful (maybe five) icons, as opposed to running through an entire menu structure. The search learns from your selections too, so if you search for "Te" often and select "Terminal", it will start showing up at the start of the search results. If you just display all of the installed applications in rows then no, it isn't faster. Structure and organization is arbitrary. Your applications aren't "structured." They are all dumped into /usr/bin or /usr/X11/bin. Somebody came along and said, "let's help people try to find these easier," and they came up with the Applications menu, which is just a set of links to the actual applications. If you think about it, it is just a "search" that has been done by somebody else, and the results have been permanently stored in .desktop files.
The search box is just an extension of the idea. It let's you perform your own arbitrary search instead of being constrained to somebody else's. Search has to be done right, obviously. Can't speak to your Windows problem because I try to avoid it as much as possible. There are a lot of good search implementations, though...GMail and Spotlight are two that come to mind immediately. I see search as supplemental to organization, not a replacement for it (although some will use it that way). I can painstakingly organize all of my files into carefully labeled folders. But search will still help me find things faster than hunting through all of my folders looking for something I last worked on a year ago.