you keep posting this same message on every MS article. Please stop
1) the timeliness of anything JVM related is impacted legally because Sun sues us for shipping a JVM and then sues us AGAIN for NOT SHIPPING a JVM. You might remark that the first time around, suing us for shipping a "doesn't play the way Sun wants us to play" JVM was perhaps justified, but then suing us AGAIN for not shipping a jvm after Sun realized that desktop java was dead because of their own stupid moves is just absurds. Seems like Sun got it both ways. The legal wrangling of java/jvm and how it infects MS is the result of that. We have to do a fire drill anytime anything about Java comes up and its all for legal reasons. I don't know for sure, but i'd suspect that there were legal reasons related to the timing of the JVM fixes.
2) The shatter attack a) has a patch that helps address it b) is not nearly as impactful as the paper makes it seem. People are making a lot bigger deal out of this than is warranted. When you know of an actual shatter based compromise that actually affects you, do let me know
3) You're basing this on the words of one person, granted, one pretty popular person in certain circles, but as there's no new patches from MS and no new SMB exploits/disclosures running around, this is neither proven nor disproven. One could argue that this was reactionary on the part of Allison, and that he didn't go back and carefully try each of those vulns he remembers coming across against W2k server or W2k3 server. You definitely don't know the answer, do you, so why do you keep bringing it up ? I mean, if I say i've found 30 security bugs in OpenBSD, but im not telling, you'd tell me i was a crack head and to sod off. How much reputation do i need to build before you extend me the same non-questioning acceptance you've given to the samba team ?
Finally, regarding the issue of patching - did you or did you NOT read the article ? Ballmer says clearly that patching is an area of known poor performance and that it is being worked on. Getting every single product at MS under a unified, smart patching system is a big undertaking, it seems reasonable that you might consider giving them a year or two to work on it. After all, how long did mac users put up with the beach ball cursor ? How long did XFree users put up with non-accelerated video ? How long did linux users put up with a joke way of handling audio hardware ? You never said "I want idiot-proof sound TOMORROW or linux is UNUSABLE!" Nobody said "mac is inherently unusauble and unfit for work becaue it has a beachball cursor that wont be fixed for over 3 weeks!"
So, please stop posting your repetitive propoganda on a consistant basis. Or at least spice it up a little with some content/insight/knowledge.
Your apache question is useless in the following sense
1) tomorrow a new webserver that _smokes_ apache comes out. oops.
2) it is asking someone to recall something memorized
tne reason some MS interview questions are so non-specific is because someone MS hires today is statistically likely to be working on a different team and different technology within 2 years. Specific knowledge like asking how to configure apache is almost useless in an environment where you have no idea what you'll be working on now or in n timeperiods from now.
some people do get very specific quesitons like that - typically, contractors are brought into to do technology-specific projects. there is no time to train them so they have to know the technologies they'll be working with on day 1. the interviews are entirely different for them because there is no built-in assumption that they have to be incredibly smart and adaptable.
If MS were going to ask a question about apache, it'd be more like this
"You have a webserver that doesn't throttle connections. Explain how you would change it so that it would. Now tell me how the adminstrator would tune or configure the throttling algorithm you've come up with"
Followup: how would it be different for an ftp server ?
Your question is mor elike the (much maligned) MCSE test. "please repeat domain-specific knowledge verbatim from some source"
The directive for controlling this is obvious to anyone that is editing the apache.conf file. The question might as well be "where is the apache conf file" because its self describingly obvious, IMO.
well, i certainly see evidence of it happening here. in seattle they're trying to give us the news paper. they call you at home, say "wed like to give you the paper". i repond " no thanks ", they follow up with "its absolutely free for time period x, we just want you to see if you like it ", i response "look, if you deliver them i'll just throw them away immediately. please dont waste your money or my time". they beleive me for a period of a few months and then try again.
the news paper form factor is ridiculous, the reporting of national/worldwide interest events is always lagged w.r.t. online sources, and finally, it seems that print reporters are universally biased against whatever it is i happen to be in support of, no matter what paper i pick up. in one particular newspaper i'd somtimes read growing up, every election day they'd print their "recommendations" on how you should vote. what kind of stalinist bullshit is that ?! that was the singular straw that broke the back for me reading newspapers.
the one big advantage print media DOES have is local coverage. i suppose i could find the online analogues of the local print papers, but realistically, its the small timer local stories that seem to upset me the most. when i read about something stupid thats close to home, its a lot more aggravating then reading about something stupid thats far away.
i highly recommend wireless lan access. i have a windows CE device with a 640x480 screen that accepts a WLAN card. When I get home and take my post-work dump, i check my "home account" email, catch up on mailing lists, and get around to reading any web type stuff i missed getting around to at work. i mean, what else are you going to do in the bathroom ?
actually the provision that was struck down was that a closed-source software purchase had to be justified. i.e. if a department wants to buy something that does w, and they opt for a non-open-source version, they have to justify that decision
its not matierally different than suggesting that you just eat table sugar. high glycemic carbs spike the blood sugar level (and cause insulin flood) similarly to taking refined sugars. insulin is the hormone of fat storage. insulin levels also inhibit glucagon, which is the fat "mover" hormone and tells (more or less) the body to suck what it needs out of fat cells to rebuild muscle, etc.
the premise of the sugar busters book is that high insulin levels are primarily caused by refined sugars and refined, high glycemic carbs. much evidence is presented that elevated insulin is one of the worst possible body conditions, leading both to pancreatic failure and insulin resistance. never mind the connections between high-insulin levels and fat uptake, and high insulin levels and "Bad cholesteral" production.
read "The Sugar Busters Diet" to get an explanation about why high glycemic carbs should be avoided at all costs. Insulin resistance is a bad, bad thing. Fast metabolizing refined sugar (and high-glycemic carbs) spike the insulin level and it remains spiked for hours. Insulin is the fat-storage hormone. Start your meal off with one poor choice (refined sugar or high-glycemic carb) and everything you eat for the next few hours gets packed into fat, essentially.
I agree with 1, 2, 3, 4, but absolutely don't agree with #5.
No. Windows has supported this since w2k Advacned server at least. Infact, i am running several 8GB Windows 2000 servers with/3GB and/PAE (They're SQL servers) NT4 enterprise edition also supported > 4GB, i beleive. W2k Datacenter went up to 32 or 64GB of ram, depending on which machine and when you got it./3GB gives processes a 3GB address space and changes the mapping for the OS components and other stuff to be in the upper 1GB range. This is really handy for apps like SQL server that can use lots of general purpose address space (especially if you run other in-proc modules, or have lots of simultaneous connections)/PAE is what lets the OS expose ram above 4GB. For an 8GB single-instance SQL box, we run the OS with/3GB and/PAE (we want SQL to have as much general purpose ram as possible (3gb addy space, and we want the OS to see all 8GB naturally:) and we enable AWE on the SQL server.
This means it has a full 3GB of address space for mapping and thread creation (there are issues iN SQL server where address space fragmentation can cause thread-create failures, effectively denying new logons. growing your address space 50% seems like it should help:) and then data/index pages can be locked into AWE pages. What this effectively means is that we can keep a lot of our database entirely in RAM, which makes perf wonderful.
maybe you weren't talking to the right people at HP. I'll just leave it at that.
Re: x86 hack. Apparently it turns out that opteron is going to run the existing base of 32 bit software really well, and MS and about every other commercial software company EVER has a huge installed base and has made lots of money off of that installed base.
Throwing it away is nonsense. Providing a smooth transition is the only reasonable thing to do. I can't beleive Intel let AMD get the 32->64 move right and screwed it up so badly themselves, at least initally.
(my opinions are based solely on being an uninformed sideliner, i have not worked with I1, I2, or AMD64 boxes personally, much less benchmarked them)
I've done interviews for both blue and orange badgers. I happen to know what the hiring/interviewing requirements are, more importantly, i know a bit about what it costs MS to bring an additional warm body online, employment wise.
In my (limited) experience, there is a totally different management attitude towards full timers and contractors. A fulltime employee is someone MS has made a HUGE financial investment in the second they sit down at their desk. Realizing the payoff of that investment takes additional, continual investment. The number of hoops you go through to complete a fulltime interview round and get a fulltime position are ridiculous, because the penalty for making a bad full time hire is enormous. Full time hires are hires for Microsoft first, they are expected to adapt to whatever project or role they are placed in and be extremely successful there. They are planned for months in advance and requests for additional headcount are scrutinized in ridiculous detail. I was hired for a team that didn't exist at the time i was interviewed. There are thousands of open positions at MS open right now that we have trouble filling because we're just that anal. To quote Ronin - "if there is any doubt, there is no doubt", is basically our approach to extending an offer to someone for fulltime employment. Not that its perfect, but we try as hard as possible to be sure someone will knock our socks off before considering them for FTE.
Contractors are entirely different. Typically a contractor interview is short, sweet, and to the point. Making a poor hire decision on a contractor costs us next to nothing compared to making a poor hire decision for FTE. The contractor is brought in because they have specific skills which are needed for a specific duration.
So from a hiring/skills perspective, FTE and contractors are looked at completely differently. The interview questions we ask are different (no relative judgement like harder/easier, just different), and the answers we expect are different to the questions that are common.
That said, most people here are humans. We've got a couple contractors on our team. I like them fine. I don't mistreat them, I don't NOT invite them to non-work sanctioned events, etc. Sometimes we bend the rules and invite them to fun events and foot the bill. It all just depends on the group and on the contractor in question.
Some contractors want to change to FTE down the road, others don't. At least one guy i knew was a FTE and decided to become a contractor so he could have 3 months of consecutive vacation every year (big snowboarder, apparently).
Some managers work with their contractors to see if they can be developed into FTE candidates (if thats what the contractor wants).
One thing we don't screw around with is sensitive information. There is a definite "need to know" attitude with contractors and disclosure on some projects. They have no vested interest in the company, and are definitely in a position to cause major competitive/financial/whatever damage. For reasons along those lines, many events/talks/whatever are FTE only.
Basically, i don't get what the article's point is. Contractors are treated differently, compensation wise, scope wise, access to information wise, and benefits wise, than full time employees. Additionally, it seems like this poster is just MS bashing, without adding anything significant to the discussion.
(par for the course, it seems, article wise and comment wise)
yes, but nobody i've ever referred has been hired, so i think im cursed in that regard:)
the microsoft.com/careers or whatever site is pretty comprehensive. my observation seems to indicate that most positions available on our internal staffing db are also posted there, so if you're seriously interested in working for "the enemy", the website is as good a place as any to start:)
i have a car with a custom subwoofer box, a 6 disc changer, a nice amp, an in-dash cd player, and so on.
i never listen to it. the previous owner of the car installed all that crap. i bought the car to drive, and when you have one of the worlds best sounding motors(*1), drowning it out with music is a crime.
who buys a car based on the factory stereo, or the music that comes in the car ? shouldn't car buying revolve around performance, safety, and price ?
yes and no. it depends which building, which campus, and which group, amongst other things. e.g. we have offices all over the world, clearly we don't solely populate every building we ever put an office in, much less own every building we ever put an office in outright.
infact, even on the redmond campus i've been told we don't _own_ all of the buildings, we lease some of them (that are apparently built to our specs), although im not aware of any main campus buildings that we share with anyone. we do have an office in downtown bellevue, wa, where we have like half of a floor or half of two floors or something. ironically enough, i think Sun also has offices in that building:)
- hotmail was an acquisition - it is advantageous from an image standpoint, and a product development standpoint, to get hotmail running on windows technologies - if you don't want to beleive what i say is credible, don't. it should be clear from my user info where i work and why i might have a marginally better idea of how this all transpired than the poster i was responding to. - regarding internal conversations - i am a nobody, but that doesn't mean i don't occasionally hear from somebodies:)
turns out that there are numerous former unix people within MS and we talk from time to time and every so often someone posts a message to an internal list saying something along the lines of "this person thinks hotmail runs unix because windows cant handle it" and then the people that _did_ the conversions chime in and explain how it really went down.
yes, the amount of FUD about hotmail is so bad that it has infected customers and even some microsoft employees (mostly field people that have to deal with said customers).
fwiw, i beleive there is some publicly available info on the current hotmail arch put out in response to all the stuff like this that people keep coming up with. searching microsoft.com might turn it up.
alternatively, you might use google to search microsoft for it:)
well, there is always terraserver. That was a > 1TB SQL database running on old windows, and old SQL server, several YEARS ago.
Look at the tpc/c scores, and at the size of some of those workloads and some of the systems. There'd be no reason to use oracle parallel server instead of MS SQL server - SQL server can handle it.
That said, yes, i highly doubt that SQL server would be the primary data store for a google-competitor. Although it would be an interesting experiment in the following sense
I unfortuneately am not an expert at windows clustering, but there are a few different ways of doing it, depending on what the task is. MSCS (MS cluster services) i beleive has a much lower node count, and requires a shared storage medium (a shared scsi cabinet or SAN volume or similar). Then theres NLB which is a distributed hash algoritm that determines which members of a host-array respond to which packets (think web farm) Typically MSCS would be used for SQL server applications, although there are also ways to break up SQL server without using any of these technologies (distributed partitioned views comes to mind)
In any case, my point was that MS _does_ have experience running some HUGE internet properties, in house, and there are other customers running other huge internet properties running on MS software. I am not claiming that MS could do google with out-of-the-box software that it has today (although i'm not saying it _couldn't_ be done, either:)
but I also know that the MS server products that are being used to run it aren't doing half the job that BSD did
how do you know that ?
is there some document that says so ? did some "reputable source" tell you ? are you just sure that BSD was doing such a better job ?
i've talked to the people involved at hotmail (and now at MS) about exactly how the transistion went. i have a good idea (from what internal documents i've read and conversations i've had) at what the old picture was and what the new picture was. i know a little bit about some of the problems and some of the successes, from essentially 1st hand sources
do you have this info ? care to share it with us all to justify your claim ?
i'll tell you what i can, and it may surprise you. the majority of freebsd front-door boxes were converted to win2k. the freebsd that was in use was essentially fully custom at that point, and not much like the stock freebsd (this would be the "special os" you mention). the w2k replacements are stock (perhaps settings tweaked) win2k servers, nothing "magic" about them. in general, the w2k boxes are outperforming the freebsd's they replaced by a statistically significant margin. not to mention other benefits (beleive it or not - reliability was one of them). note that this is on the same hardware - freebsd was replaced with w2k
some people will say "netcraft still says its freebsd!" and they're partially correct. there are still some freebsd boxes that remain at hotmail last time i talked to anybody about it. additionally, the Back end mailstore machines were neither windows or freebsd - they were giant sun boxes that were already purchased and too expensive to simply throw away only on the gorunds of doing it all "the windows way"... although i understand that converting them is now underway or coming soon..
initally, i didn't beleive it either when i read the documents. turns out the hotmail conversion really shaped a lot of w2k because they started working on it early. w2k is partially what it is today because it had to exceed freebsd enough to make the conversion not only possible, but worthwhile.
you should keep a few things in mind when making blanket statements about what windows can and cant do
microsoft.com runs windows. it's a website that gets more traffic than most. it stands up
microsoft, as a company and as a network entity, gets hit with more attacks than just about any other corporate or network presence. generally, it holds up
sql servers running on windows are regularly trading tpc/c 1st place scores with other machines - we're now in the realm of doing all of the NYSE transactions for a YEAR in a matter of a day or so
i use google exclusively for my searching. i use google to search microsoft sites. I use google to help me search MSDN. The people that write MSDN work in the same building as I do.
I think google is the stuff and i rarely see a need to use anything else. it is -Exactly- the interface that i want and it is lightning fast with no distractions.
perhaps there is some breakthough in searching/indexing technology that MS thinks they can make. I'd buy that - there are lots of bright people here that really understand interesting problems with other approaches and can turn that understanding into solid products.
or perhaps someone decided google isn't "friendly" enough (i.e. not filled with crap, ads, marketing tie-ins, etc etc) and needs to be cluttered up and "popularized". maybe someone simply wants MS to have the #1 search and thats driving the whole story..
My worry is that whatever comes out of this, it will end up being 800kb of dhtml and popups and shitty ads. I don't think anything will ever replace google for what the majority of people use it for unless it is as simple and stripped down as google is, interface wise. i mean, i have a vested financial interest in MS products doing well but i still find myself using what i feel is the right tool for the job which fits my usage habits best, and for basically all searching tasks thats google.
you can't have it both ways. If the contention is that MCSE's are idiots and unix admins are smart, then i get to say things like "unix admins are required to be smart"
which is it ? the competancy required to manage and use unix and windows is the same ? or are unix users/admins "more capable" than windows users/admins ?
if you buy that unix users and admins are brighter than windows/users and admins, is it because englightened people always choose the same thing (hardly), or is it out of necessity, (you wont see a dumb unix admin because you cant be dumb and admin unix). i think it's easy to make the claim that the average UNIX guy is considered a guru amongst his non-unix employees. if i had left out "elitist" would you have been happier ?
Re:Ha, ha. You joke. Is good.
on
BSA IDC FUD
·
· Score: 1
there are two issues with your comment
1) criteria for a fix 2) quality of the fix
1) developers will only fix things that affect them (or that they notice, or are paid to fix)
2) developers will only fix things to the satisfaction of themselves
taken together, this means that my grandmother wont get fixes that apply to her problems, and the fixes she may get will be things like "recompile the kernel first, use xvidtune" and so on.
yes, it is possible to apply quickly developed ramshackle source-patches to a box and it will get fixed quick. That doesn't help anyone except the person administering that box. That helps the adminstrator's job security immensely hwoever, as that person can rightfully say "only I know how this box was built and only I know how to keep it running"
Surely you've come across someone elses unix machine and thought "what the hell are they thinking ?" In a system peiced together from source, how is the end result substantially different from coming into ownership of some other developers source code ? Have you experienced trying to take ownership of someone elses source ? Is this a realistic approach to managing business systems ? That to run a web server, the admin must pass his or her tome of knowledge onto the next person that will run it ?
My contention is that the major cost of IT is human labour. OSS does little to reduce that; on the contrary, my point is that the idealogy and practice related to many OSS projects are such that they promote individual ownership and the "we're fucked if she gets hit by a bus" phenomena, thus making people extremely expensive and difficult to replace a single person.
naturally IT workers will promote OSS and these types of projects as much as possible. "spend nothing on the software so you can pay me MORE to run it, AND let me put you in a position where its nearly impossible to fire me!"
Re:Easy...
on
BSA IDC FUD
·
· Score: 3, Interesting
actually, you're dead on.
Open source software _is_ good for the IT industry. Broken software that requires babysitting by elitist gurus is _exactly_ what IT workers want, so they can continue to justify their positions and their salaries.
UNIX and Open source in general are _Great_ for the privileged few IT workers that use them effectively (or use them effectively enough to fool their employers).
Until companies start doing the hard analysis of "gosh, even though i sell shoes, IT is 50% of my expenditures. Maybe i should go back to the old way and cut my costs, after all, any 5.75/hr secretary can file papers and write order tickets"
Then IT industry will crash and the people that had cushy jobs because they were pseudo-wizards will get laid off, and companies will start using software that doesn't require wizards to run, and actually lets them focus on their business instead of their IT dept.
Not that any UNIX/internet companies have had trouble or layoffs recently, or anything;)
because i had normal mod points when i saw this, but unfortuneately the story submitter is an IDIOT, and michael is also an IDIOT, and many of the people posting responses haven't read ANYTHING related to the article except the posted blurb by the first IDIOT, and thus look like IDIOTS themselves.
The slashdot submissions clearly says that microsoft and the MPAA are both testifying that piracy supports terrorism.
"[Yesterday's] Oversight Hearing on "International Copyright Piracy: Links to Organized Crime and Terrorism" featured the MPAA and Microsoft testifying that software and movie DVD counterfeiting is an acute problem, with criminal gangs operating factories in Russia, Malaysia and other countries that have weak copyright laws. They further claim that intellectual property piracy is a vehicle for financing or supporting acts of terror."
BULLSHIT
http://www.house.gov/judiciary/lamagna031303.pdf
Here is the exact testimony of the microsoft lawyer. Terrorism is not mentioned a single time.
Microsoft's only contention here is that the majority of large scale piracy is done by very well funded operations with links to organized crime, primarily backed by and operating in countries with less strict or non-existant IP laws. It then goes on to say that much of the profit (and its nearly ALL profit) of these operations goes to funding other activity within those crime organizations, some of which is violent crime. There is PROOF of this cited in the comments. The only part of it that is conjecture is the estimated revenue and job losses due to piracy, the arguments against which are well known and do not need to be repeated here.
Nowhere in the microsoft testimony, nor in the ZDNET article is there any link between MS testimony and terrorism _at all_. Nowhere is MS claiming that piracy causes terrorism. Nowhere is there anything to indicate that MS and the MPAA are best friends in crushing your inner child.
This website might as well change its name to "microsoft_enquierer" or "microsoftdailysun" or some similar such tabloid name.
Oh wait! we already have theregister (which nearly every MS related article on slashdot invariably links to as an authoritative or credible source of "journalism")
If slashdot is going to try and act as a political or any other kind of entity, stick to the facts, clearly differentiate conjecture from reality, and at least make a half hearted attempt at being accurate.
please explain where the threat is that's being downplayed ?
If this were the worst "issue" with Windows security, nobody would use anything else. Nobody.
In my opinion, this issue isn't on the panic scale at all - it's on the "everyone that's worried about it is a fuckwit" scale, weighing in right around 9.5.
This article has nothing to do with being a windows apologist. This issue effects essentially ALL pc operating systems. Just last week i floppy-booted my openBSD machine because i forgot the root password, then changed it. Where is the media frenzy ? Where, as another posted pointed out, is the get-theo-lynchmob ?
AFAIK it's easier to totally circumvent ANY pc unix machine with a bootfloppy (unless its configured specially) than it is to use this recovery console trick to do anything of gain to a windows machine. what a fantastic showing by microsoft if this is worthy of harassment.. because the message is "you handle something marginally better than all other currently widely used OSes"
Listen, this is slashdot. It's ok (expected, even!) to hate windows and microsoft. But to be really effective, you should pick something worth being angry about. And if you can't find anything better to get concerned with than this, you really don't have much justification for concern at all (and you don't have much of a justification to comment on the matter, either)
why would you think that MS is sucking up to the RIAA ?
If the INDUSTRY (thats MS, Intel, and others) dont create the technology specification for things like TCPA and Palladium, then the RIAA will and will make circumventing it illegal.
Which do you want ? Technology solutions without legal authority, or bullshit that comes with the requirement of federal prosecution and jailtime ?
Put another way - MS doesn't want the RIAA's shit in its software.
(And as much as anyone dislikes MS, surely you dislike the RIAA _more_) (And as bad as you think MS software is, have you USED the stuff put out by RIAA types ? look at what happened to the lyrics websites with that TERRIBLE java applet you have to use now...)
Slashdot Mirror
you keep posting this same message on every MS article. Please stop
1) the timeliness of anything JVM related is impacted legally because Sun sues us for shipping a JVM and then sues us AGAIN for NOT SHIPPING a JVM. You might remark that the first time around, suing us for shipping a "doesn't play the way Sun wants us to play" JVM was perhaps justified, but then suing us AGAIN for not shipping a jvm after Sun realized that desktop java was dead because of their own stupid moves is just absurds. Seems like Sun got it both ways. The legal wrangling of java/jvm and how it infects MS is the result of that. We have to do a fire drill anytime anything about Java comes up and its all for legal reasons. I don't know for sure, but i'd suspect that there were legal reasons related to the timing of the JVM fixes.
2) The shatter attack a) has a patch that helps address it b) is not nearly as impactful as the paper makes it seem. People are making a lot bigger deal out of this than is warranted. When you know of an actual shatter based compromise that actually affects you, do let me know
3) You're basing this on the words of one person, granted, one pretty popular person in certain circles, but as there's no new patches from MS and no new SMB exploits/disclosures running around, this is neither proven nor disproven. One could argue that this was reactionary on the part of Allison, and that he didn't go back and carefully try each of those vulns he remembers coming across against W2k server or W2k3 server. You definitely don't know the answer, do you, so why do you keep bringing it up ? I mean, if I say i've found 30 security bugs in OpenBSD, but im not telling, you'd tell me i was a crack head and to sod off. How much reputation do i need to build before you extend me the same non-questioning acceptance you've given to the samba team ?
Finally, regarding the issue of patching - did you or did you NOT read the article ? Ballmer says clearly that patching is an area of known poor performance and that it is being worked on. Getting every single product at MS under a unified, smart patching system is a big undertaking, it seems reasonable that you might consider giving them a year or two to work on it. After all, how long did mac users put up with the beach ball cursor ? How long did XFree users put up with non-accelerated video ? How long did linux users put up with a joke way of handling audio hardware ? You never said "I want idiot-proof sound TOMORROW or linux is UNUSABLE!" Nobody said "mac is inherently unusauble and unfit for work becaue it has a beachball cursor that wont be fixed for over 3 weeks!"
So, please stop posting your repetitive propoganda on a consistant basis. Or at least spice it up a little with some content/insight/knowledge.
Your apache question is useless in the following sense
1) tomorrow a new webserver that _smokes_ apache comes out. oops.
2) it is asking someone to recall something memorized
tne reason some MS interview questions are so non-specific is because someone MS hires today is statistically likely to be working on a different team and different technology within 2 years. Specific knowledge like asking how to configure apache is almost useless in an environment where you have no idea what you'll be working on now or in n timeperiods from now.
some people do get very specific quesitons like that - typically, contractors are brought into to do technology-specific projects. there is no time to train them so they have to know the technologies they'll be working with on day 1. the interviews are entirely different for them because there is no built-in assumption that they have to be incredibly smart and adaptable.
If MS were going to ask a question about apache, it'd be more like this
"You have a webserver that doesn't throttle connections. Explain how you would change it so that it would. Now tell me how the adminstrator would tune or configure the throttling algorithm you've come up with"
Followup: how would it be different for an ftp server ?
Your question is mor elike the (much maligned) MCSE test. "please repeat domain-specific knowledge verbatim from some source"
The directive for controlling this is obvious to anyone that is editing the apache.conf file. The question might as well be "where is the apache conf file" because its self describingly obvious, IMO.
well, i certainly see evidence of it happening here. in seattle they're trying to give us the news paper. they call you at home, say "wed like to give you the paper". i repond " no thanks ", they follow up with "its absolutely free for time period x, we just want you to see if you like it ", i response "look, if you deliver them i'll just throw them away immediately. please dont waste your money or my time". they beleive me for a period of a few months and then try again.
the news paper form factor is ridiculous, the reporting of national/worldwide interest events is always lagged w.r.t. online sources, and finally, it seems that print reporters are universally biased against whatever it is i happen to be in support of, no matter what paper i pick up. in one particular newspaper i'd somtimes read growing up, every election day they'd print their "recommendations" on how you should vote. what kind of stalinist bullshit is that ?! that was the singular straw that broke the back for me reading newspapers.
the one big advantage print media DOES have is local coverage. i suppose i could find the online analogues of the local print papers, but realistically, its the small timer local stories that seem to upset me the most. when i read about something stupid thats close to home, its a lot more aggravating then reading about something stupid thats far away.
i highly recommend wireless lan access. i have a windows CE device with a 640x480 screen that accepts a WLAN card. When I get home and take my post-work dump, i check my "home account" email, catch up on mailing lists, and get around to reading any web type stuff i missed getting around to at work. i mean, what else are you going to do in the bathroom ?
actually the provision that was struck down was that a closed-source software purchase had to be justified. i.e. if a department wants to buy something that does w, and they opt for a non-open-source version, they have to justify that decision
its not matierally different than suggesting that you just eat table sugar. high glycemic carbs spike the blood sugar level (and cause insulin flood) similarly to taking refined sugars. insulin is the hormone of fat storage. insulin levels also inhibit glucagon, which is the fat "mover" hormone and tells (more or less) the body to suck what it needs out of fat cells to rebuild muscle, etc.
the premise of the sugar busters book is that high insulin levels are primarily caused by refined sugars and refined, high glycemic carbs. much evidence is presented that elevated insulin is one of the worst possible body conditions, leading both to pancreatic failure and insulin resistance. never mind the connections between high-insulin levels and fat uptake, and high insulin levels and "Bad cholesteral" production.
read "The Sugar Busters Diet" to get an explanation about why high glycemic carbs should be avoided at all costs. Insulin resistance is a bad, bad thing. Fast metabolizing refined sugar (and high-glycemic carbs) spike the insulin level and it remains spiked for hours. Insulin is the fat-storage hormone. Start your meal off with one poor choice (refined sugar or high-glycemic carb) and everything you eat for the next few hours gets packed into fat, essentially.
I agree with 1, 2, 3, 4, but absolutely don't agree with #5.
No. Windows has supported this since w2k Advacned server at least. Infact, i am running several 8GB Windows 2000 servers with /3GB and /PAE (They're SQL servers) NT4 enterprise edition also supported > 4GB, i beleive. W2k Datacenter went up to 32 or 64GB of ram, depending on which machine and when you got it. /3GB gives processes a 3GB address space and changes the mapping for the OS components and other stuff to be in the upper 1GB range. This is really handy for apps like SQL server that can use lots of general purpose address space (especially if you run other in-proc modules, or have lots of simultaneous connections) /PAE is what lets the OS expose ram above 4GB. For an 8GB single-instance SQL box, we run the OS with /3GB and /PAE (we want SQL to have as much general purpose ram as possible (3gb addy space, and we want the OS to see all 8GB naturally :) and we enable AWE on the SQL server.
:) and then data/index pages can be locked into AWE pages. What this effectively means is that we can keep a lot of our database entirely in RAM, which makes perf wonderful.
This means it has a full 3GB of address space for mapping and thread creation (there are issues iN SQL server where address space fragmentation can cause thread-create failures, effectively denying new logons. growing your address space 50% seems like it should help
maybe you weren't talking to the right people at HP. I'll just leave it at that.
Re: x86 hack. Apparently it turns out that opteron is going to run the existing base of 32 bit software really well, and MS and about every other commercial software company EVER has a huge installed base and has made lots of money off of that installed base.
Throwing it away is nonsense. Providing a smooth transition is the only reasonable thing to do. I can't beleive Intel let AMD get the 32->64 move right and screwed it up so badly themselves, at least initally.
(my opinions are based solely on being an uninformed sideliner, i have not worked with I1, I2, or AMD64 boxes personally, much less benchmarked them)
I'm a blue badger.
I've done interviews for both blue and orange badgers. I happen to know what the hiring/interviewing requirements are, more importantly, i know a bit about what it costs MS to bring an additional warm body online, employment wise.
In my (limited) experience, there is a totally different management attitude towards full timers and contractors. A fulltime employee is someone MS has made a HUGE financial investment in the second they sit down at their desk. Realizing the payoff of that investment takes additional, continual investment. The number of hoops you go through to complete a fulltime interview round and get a fulltime position are ridiculous, because the penalty for making a bad full time hire is enormous. Full time hires are hires for Microsoft first, they are expected to adapt to whatever project or role they are placed in and be extremely successful there. They are planned for months in advance and requests for additional headcount are scrutinized in ridiculous detail. I was hired for a team that didn't exist at the time i was interviewed. There are thousands of open positions at MS open right now that we have trouble filling because we're just that anal. To quote Ronin - "if there is any doubt, there is no doubt", is basically our approach to extending an offer to someone for fulltime employment. Not that its perfect, but we try as hard as possible to be sure someone will knock our socks off before considering them for FTE.
Contractors are entirely different. Typically a contractor interview is short, sweet, and to the point. Making a poor hire decision on a contractor costs us next to nothing compared to making a poor hire decision for FTE. The contractor is brought in because they have specific skills which are needed for a specific duration.
So from a hiring/skills perspective, FTE and contractors are looked at completely differently. The interview questions we ask are different (no relative judgement like harder/easier, just different), and the answers we expect are different to the questions that are common.
That said, most people here are humans. We've got a couple contractors on our team. I like them fine. I don't mistreat them, I don't NOT invite them to non-work sanctioned events, etc. Sometimes we bend the rules and invite them to fun events and foot the bill. It all just depends on the group and on the contractor in question.
Some contractors want to change to FTE down the road, others don't. At least one guy i knew was a FTE and decided to become a contractor so he could have 3 months of consecutive vacation every year (big snowboarder, apparently).
Some managers work with their contractors to see if they can be developed into FTE candidates (if thats what the contractor wants).
One thing we don't screw around with is sensitive information. There is a definite "need to know" attitude with contractors and disclosure on some projects. They have no vested interest in the company, and are definitely in a position to cause major competitive/financial/whatever damage. For reasons along those lines, many events/talks/whatever are FTE only.
Basically, i don't get what the article's point is. Contractors are treated differently, compensation wise, scope wise, access to information wise, and benefits wise, than full time employees. Additionally, it seems like this poster is just MS bashing, without adding anything significant to the discussion.
(par for the course, it seems, article wise and comment wise)
yes, but nobody i've ever referred has been hired, so i think im cursed in that regard :)
:)
the microsoft.com/careers or whatever site is pretty comprehensive. my observation seems to indicate that most positions available on our internal staffing db are also posted there, so if you're seriously interested in working for "the enemy", the website is as good a place as any to start
E28 M5 :)
:)
Congrats on either knowing your stuff, or being effective with google
i have a car with a custom subwoofer box, a 6 disc changer, a nice amp, an in-dash cd player, and so on.
i never listen to it. the previous owner of the car installed all that crap. i bought the car to drive, and when you have one of the worlds best sounding motors(*1), drowning it out with music is a crime.
who buys a car based on the factory stereo, or the music that comes in the car ? shouldn't car buying revolve around performance, safety, and price ?
*1- S38B35 US spec, with catalyst
yes and no. it depends which building, which campus, and which group, amongst other things. e.g. we have offices all over the world, clearly we don't solely populate every building we ever put an office in, much less own every building we ever put an office in outright.
:)
infact, even on the redmond campus i've been told we don't _own_ all of the buildings, we lease some of them (that are apparently built to our specs), although im not aware of any main campus buildings that we share with anyone. we do have an office in downtown bellevue, wa, where we have like half of a floor or half of two floors or something. ironically enough, i think Sun also has offices in that building
a few things to understand:
:)
:)
- hotmail was an acquisition
- it is advantageous from an image standpoint, and a product development standpoint, to get hotmail running on windows technologies
- if you don't want to beleive what i say is credible, don't. it should be clear from my user info where i work and why i might have a marginally better idea of how this all transpired than the poster i was responding to.
- regarding internal conversations - i am a nobody, but that doesn't mean i don't occasionally hear from somebodies
turns out that there are numerous former unix people within MS and we talk from time to time and every so often someone posts a message to an internal list saying something along the lines of "this person thinks hotmail runs unix because windows cant handle it" and then the people that _did_ the conversions chime in and explain how it really went down.
yes, the amount of FUD about hotmail is so bad that it has infected customers and even some microsoft employees (mostly field people that have to deal with said customers).
fwiw, i beleive there is some publicly available info on the current hotmail arch put out in response to all the stuff like this that people keep coming up with. searching microsoft.com might turn it up.
alternatively, you might use google to search microsoft for it
well, there is always terraserver. That was a > 1TB SQL database running on old windows, and old SQL server, several YEARS ago.
:)
Look at the tpc/c scores, and at the size of some of those workloads and some of the systems. There'd be no reason to use oracle parallel server instead of MS SQL server - SQL server can handle it.
That said, yes, i highly doubt that SQL server would be the primary data store for a google-competitor. Although it would be an interesting experiment in the following sense
I unfortuneately am not an expert at windows clustering, but there are a few different ways of doing it, depending on what the task is. MSCS (MS cluster services) i beleive has a much lower node count, and requires a shared storage medium (a shared scsi cabinet or SAN volume or similar). Then theres NLB which is a distributed hash algoritm that determines which members of a host-array respond to which packets (think web farm) Typically MSCS would be used for SQL server applications, although there are also ways to break up SQL server without using any of these technologies (distributed partitioned views comes to mind)
In any case, my point was that MS _does_ have experience running some HUGE internet properties, in house, and there are other customers running other huge internet properties running on MS software. I am not claiming that MS could do google with out-of-the-box software that it has today (although i'm not saying it _couldn't_ be done, either
Read my user info :)
(hint: Yes I do)
how do you know that ?
is there some document that says so ? did some "reputable source" tell you ? are you just sure that BSD was doing such a better job ?
i've talked to the people involved at hotmail (and now at MS) about exactly how the transistion went. i have a good idea (from what internal documents i've read and conversations i've had) at what the old picture was and what the new picture was. i know a little bit about some of the problems and some of the successes, from essentially 1st hand sources
do you have this info ? care to share it with us all to justify your claim ?
i'll tell you what i can, and it may surprise you. the majority of freebsd front-door boxes were converted to win2k. the freebsd that was in use was essentially fully custom at that point, and not much like the stock freebsd (this would be the "special os" you mention). the w2k replacements are stock (perhaps settings tweaked) win2k servers, nothing "magic" about them. in general, the w2k boxes are outperforming the freebsd's they replaced by a statistically significant margin. not to mention other benefits (beleive it or not - reliability was one of them). note that this is on the same hardware - freebsd was replaced with w2k
some people will say "netcraft still says its freebsd!" and they're partially correct. there are still some freebsd boxes that remain at hotmail last time i talked to anybody about it. additionally, the Back end mailstore machines were neither windows or freebsd - they were giant sun boxes that were already purchased and too expensive to simply throw away only on the gorunds of doing it all "the windows way"... although i understand that converting them is now underway or coming soon..
initally, i didn't beleive it either when i read the documents. turns out the hotmail conversion really shaped a lot of w2k because they started working on it early. w2k is partially what it is today because it had to exceed freebsd enough to make the conversion not only possible, but worthwhile.
you should keep a few things in mind when making blanket statements about what windows can and cant do
i use google exclusively for my searching. i use google to search microsoft sites. I use google to help me search MSDN. The people that write MSDN work in the same building as I do.
I think google is the stuff and i rarely see a need to use anything else. it is -Exactly- the interface that i want and it is lightning fast with no distractions.
perhaps there is some breakthough in searching/indexing technology that MS thinks they can make. I'd buy that - there are lots of bright people here that really understand interesting problems with other approaches and can turn that understanding into solid products.
or perhaps someone decided google isn't "friendly" enough (i.e. not filled with crap, ads, marketing tie-ins, etc etc) and needs to be cluttered up and "popularized". maybe someone simply wants MS to have the #1 search and thats driving the whole story..
My worry is that whatever comes out of this, it will end up being 800kb of dhtml and popups and shitty ads. I don't think anything will ever replace google for what the majority of people use it for unless it is as simple and stripped down as google is, interface wise. i mean, i have a vested financial interest in MS products doing well but i still find myself using what i feel is the right tool for the job which fits my usage habits best, and for basically all searching tasks thats google.
can your dad configure a unix mail reader ?
can your dad setup apache ?
you can't have it both ways. If the contention is that MCSE's are idiots and unix admins are smart, then i get to say things like "unix admins are required to be smart"
which is it ? the competancy required to manage and use unix and windows is the same ? or are unix users/admins "more capable" than windows users/admins ?
if you buy that unix users and admins are brighter than windows/users and admins, is it because englightened people always choose the same thing (hardly), or is it out of necessity, (you wont see a dumb unix admin because you cant be dumb and admin unix). i think it's easy to make the claim that the average UNIX guy is considered a guru amongst his non-unix employees. if i had left out "elitist" would you have been happier ?
there are two issues with your comment
1) criteria for a fix
2) quality of the fix
1) developers will only fix things that affect them (or that they notice, or are paid to fix)
2) developers will only fix things to the satisfaction of themselves
taken together, this means that my grandmother wont get fixes that apply to her problems, and the fixes she may get will be things like "recompile the kernel first, use xvidtune" and so on.
yes, it is possible to apply quickly developed ramshackle source-patches to a box and it will get fixed quick. That doesn't help anyone except the person administering that box. That helps the adminstrator's job security immensely hwoever, as that person can rightfully say "only I know how this box was built and only I know how to keep it running"
Surely you've come across someone elses unix machine and thought "what the hell are they thinking ?" In a system peiced together from source, how is the end result substantially different from coming into ownership of some other developers source code ? Have you experienced trying to take ownership of someone elses source ? Is this a realistic approach to managing business systems ? That to run a web server, the admin must pass his or her tome of knowledge onto the next person that will run it ?
My contention is that the major cost of IT is human labour. OSS does little to reduce that; on the contrary, my point is that the idealogy and practice related to many OSS projects are such that they promote individual ownership and the "we're fucked if she gets hit by a bus" phenomena, thus making people extremely expensive and difficult to replace a single person.
naturally IT workers will promote OSS and these types of projects as much as possible. "spend nothing on the software so you can pay me MORE to run it, AND let me put you in a position where its nearly impossible to fire me!"
actually, you're dead on.
;)
Open source software _is_ good for the IT industry. Broken software that requires babysitting by elitist gurus is _exactly_ what IT workers want, so they can continue to justify their positions and their salaries.
UNIX and Open source in general are _Great_ for the privileged few IT workers that use them effectively (or use them effectively enough to fool their employers).
Until companies start doing the hard analysis of "gosh, even though i sell shoes, IT is 50% of my expenditures. Maybe i should go back to the old way and cut my costs, after all, any 5.75/hr secretary can file papers and write order tickets"
Then IT industry will crash and the people that had cushy jobs because they were pseudo-wizards will get laid off, and companies will start using software that doesn't require wizards to run, and actually lets them focus on their business instead of their IT dept.
Not that any UNIX/internet companies have had trouble or layoffs recently, or anything
because i had normal mod points when i saw this, but unfortuneately the story submitter is an IDIOT, and michael is also an IDIOT, and many of the people posting responses haven't read ANYTHING related to the article except the posted blurb by the first IDIOT, and thus look like IDIOTS themselves.
f
The slashdot submissions clearly says that microsoft and the MPAA are both testifying that piracy supports terrorism.
"[Yesterday's] Oversight Hearing on "International Copyright Piracy: Links to Organized Crime and Terrorism" featured the MPAA and Microsoft testifying that software and movie DVD counterfeiting is an acute problem, with criminal gangs operating factories in Russia, Malaysia and other countries that have weak copyright laws. They further claim that intellectual property piracy is a vehicle for financing or supporting acts of terror."
BULLSHIT
http://www.house.gov/judiciary/lamagna031303.pd
Here is the exact testimony of the microsoft lawyer. Terrorism is not mentioned a single time.
Microsoft's only contention here is that the majority of large scale piracy is done by very well funded operations with links to organized crime, primarily backed by and operating in countries with less strict or non-existant IP laws. It then goes on to say that much of the profit (and its nearly ALL profit) of these operations goes to funding other activity within those crime organizations, some of which is violent crime. There is PROOF of this cited in the comments. The only part of it that is conjecture is the estimated revenue and job losses due to piracy, the arguments against which are well known and do not need to be repeated here.
Nowhere in the microsoft testimony, nor in the ZDNET article is there any link between MS testimony and terrorism _at all_. Nowhere is MS claiming that piracy causes terrorism. Nowhere is there anything to indicate that MS and the MPAA are best friends in crushing your inner child.
This website might as well change its name to "microsoft_enquierer" or "microsoftdailysun" or some similar such tabloid name.
Oh wait! we already have theregister (which nearly every MS related article on slashdot invariably links to as an authoritative or credible source of "journalism")
If slashdot is going to try and act as a political or any other kind of entity, stick to the facts, clearly differentiate conjecture from reality, and at least make a half hearted attempt at being accurate.
please explain where the threat is that's being downplayed ?
If this were the worst "issue" with Windows security, nobody would use anything else. Nobody.
In my opinion, this issue isn't on the panic scale at all - it's on the "everyone that's worried about it is a fuckwit" scale, weighing in right around 9.5.
This article has nothing to do with being a windows apologist. This issue effects essentially ALL pc operating systems. Just last week i floppy-booted my openBSD machine because i forgot the root password, then changed it. Where is the media frenzy ? Where, as another posted pointed out, is the get-theo-lynchmob ?
AFAIK it's easier to totally circumvent ANY pc unix machine with a bootfloppy (unless its configured specially) than it is to use this recovery console trick to do anything of gain to a windows machine. what a fantastic showing by microsoft if this is worthy of harassment.. because the message is "you handle something marginally better than all other currently widely used OSes"
Listen, this is slashdot. It's ok (expected, even!) to hate windows and microsoft. But to be really effective, you should pick something worth being angry about. And if you can't find anything better to get concerned with than this, you really don't have much justification for concern at all (and you don't have much of a justification to comment on the matter, either)
why would you think that MS is sucking up to the RIAA ?
If the INDUSTRY (thats MS, Intel, and others) dont create the technology specification for things like TCPA and Palladium, then the RIAA will and will make circumventing it illegal.
Which do you want ? Technology solutions without legal authority, or bullshit that comes with the requirement of federal prosecution and jailtime ?
Put another way - MS doesn't want the RIAA's shit in its software.
(And as much as anyone dislikes MS, surely you dislike the RIAA _more_)
(And as bad as you think MS software is, have you USED the stuff put out by RIAA types ? look at what happened to the lyrics websites with that TERRIBLE java applet you have to use now...)