The problem is that the "maximum you are willing to pay" assumption is flawed.
I don't know the maximum I am willing to pay, because I don't know the "value" of an item. Why don't I know the value of the item? Because I don't know what other people are willing to pay for it. What the marketplace is willing to pay for something determines its value. I can't choose a bid ahead of time that will guarantee that I win the auctino yet still guarantee that I pay something I can afford.
If my max bid is in the ballpark of what I'll spend, I'll still spend 1% extra to outbid someone else, thus blowing my "maximum".
I chould choose, $20, $80, or $200 as a max bid for something. But at any of those price points, I'd be willing to pay $21, $81, or $201, because it represents such a trivial incremental addition to the price. If i want the thing bad enough to bid on it, I want to win it. Quibbling over a few dollars is less important than actually winning.
After I lost a few Cisco 678 auctions, I just sniped the next one that came up (and paid less, but that was not my motivation for doing it)
I am a big fan of the _idea_ of public transit.. I spent a few weeks in Munich and it was wonderful..but munich has, iirc, 9 u-bahn and 27 s-bahn lines for a metro area of about 1 million people. Isn't Marienplatz or Munich Hauptbanhof 4-5 levels deep ? I have no idea how a population of 1m people can support such an incredible public train system (but I sure enjoyed it while I was there!).
When we went to Berlin (via car - we toured Germany via car and I also had a quick stop to drive 6 laps of the Nordschleife) we just parked at the most distant P&R we could fine. Touring any large German city via car is simply pointless. The U-bahn system in berlin was very poor compared to Munich, but then, they're still re-assembling things in light of the 1989 re-unification.. and their system is much older and has more legacy-inspired problems afaict.
Now, given that Germany has excellent mass transit in dense cities... but also the most excellent highway system on the planet.. doesn't it seem like good public transit in beautiful dense cities and excellent highway systems are not mutually exclusive? Fwiw, Germany also has the ICE rail system which is frankly faster than any car you can rent easily.
Now, onto the US.
Eisenhower's goal was military, with the civilian benefits being just that - benefits. Next time you're scooting along the interstate in rural america.. going 70-80mph, imagine trying to go even 1/3rd that speed on the ground even 10feet away from the edge of the road.. it doesn't matter what kind of vehicle we're talking about; moving over unimproved ground at any kind of speed is near impossible, and certainly hard on equipment and people.
Christopher Alexander points out many of the shortcomings you do w.r.t. our car-focused society. My wife and I live "downtown" and we have to drive out to the subburbs/strip malls to buy groceries (although last evening we bicycled). It was after our Germany trip that we realized how good it feels to walk places instead of just hopping in the car. But you (and Alexander) point out many of the defects in American city/road planning that cause them to be car-optimized and pedestrian antagonistic.
It's hard to say that Ike made a "mistake". It's not a forgone conclusion that the US would have developed identically to europe even without our interstate system. And it's also not a forgone conclusion that the US would not have suffered some of the same ills without an interstate system. For instance, though the smog in LA is bad, isn't the air in London pretty bad also? And don't they have a pretty extensive underground rail system? Presumably, fewer Londoners depend on personal automobiles for daily commuting yet the air is still (reputedly) poor.
FAT32 and NTFS, are ages behind file systems like ResierFS (especially reiserfs v4) and even Ext3
Beleive me, I'm no huge fan of NTFS, but could you explain why Ext3 is ages ahead of it?
There's a lot of interesting work going on Reiser - that I'll certainly agree with. Although if anecdotes about people losing data with it are any indication, "stability" isn't one of the advantages it has over NTFS (at least for now..)
The only people that have been in the US for more than about 30 generations are laregly confined to "reservations", which are hotbeds of crime, inebriation, and a dead-end life.
(at least, that's my impression from visiting a few reservations, including the tribe of which my great grandmother was 100% full blooded and where I probably have a few distant relatives...)
The idea that you are "more american" because your family has been here for "several generations" is dubious. How many generations must have been born here for you to be a "real American" ?
In my opinion, anyone that is willing to follow the laws of this country should be able to become a citizen, easily and cheaply. If you think your birth pedigree is of relevance in what your rights should be, you are hardly living in the spirt of the principles on which America was founded.. perhaps you'd be happier in colonial/monarchial Europe of a few hundred years ago, or India under the caste system?
The immigration policy of this country should be guided by the following principle: "If you are willing to come here legally and comply with our laws, stay as long as you like"
That said, depending on who you are talking to, some employees will say "use the MS product in space blah no mater what.. because if you don't think it's the best for you, you need to help us understand how to make it better"
Other employees say "sometimes I don't want to be Microsoft's First Customer - I just want to do my thing without any hassles"
Very few employees are 100% MSFT all the time for all topics. Google _is_ a popular search destination. The issue of Google vs MSN/windows live search is a sore one for lots of employees. The guys behind the search technologies have setup some clever multi-search stuff internally where you can vote on results from the various different engines (which does account for some fraction of any microsoft-hitting-google traffic, btw) and then they decide how to roll human selectivity feedback into the technology. They'll tell everyone to use our stuff, even if its via one of the multi-search proxies, because they want all the data/criticism they can get to try and make it work better. The employees that just use google aren't thrilled either - tmore often than not, hey'd _like_ to use our stuff, but they just dont have the patience to try something else first, then get frustrated with the results, THEN go to Google.
In my years at Microsoft, I have never seen any evidence to indicate that we have any kind of outbound filteirng whatsoever. Certainly there is proxying and logging, but I've never seen anything blocked. I think the policy must be something like "we assume our employees are reasonable, intelligent people. We can give them free reign and if they abuse that priviledge we have logs to support any disciplinary action we may take"
Also, for essentially any content you can think of, it is somebody'sjob to look at that content. (eg the IE test team needs to make sure google renders properly...maybe people in our MSN properties need to be able to look at "inappropriate" material at times as part of cooperation with FBI investigations.. our developers are not allowed to look at online patent docs or GPL source code (for fear of IP pollution lawsuits hitting us), but our legal team HAS to in order to do some of thier job...)
Figuring out what should be filtered for what people would be a lot of effort. It's probably easier to just fire people that do things they shouldn't:)
I was thinking of posting something like this. Something like "I am sure the Chinese spec has monitoring backdoors in it known only to the Chinese govt". I'd say this because I'd be implying that whatever the US is in favor of wouldn't have such drawbacks.
Except I don't think it's reasonable to think or say such a thing in light of recent events in the US, which is a shame.
I was going to ask the same question. A friend of mine works on adserving technology and i was giving him a bit of greif about how he slept at nite, and so on.
He and I are both car junkies so he had a clever response. "If, when you saw ads, they were things like new products for your specific car, would you be as mad at them? I mean, if someone makes a new 9lb flywheel for your engine, and we show you that ad, will it be upsetting?"
I had to concede - no. I currently spend my time trying to find what I want when it comes to go-fast parts for my cars.
If I only ever saw ads for performance car parts for cars that I own, deals on new anime releases, and accessories for canon EOS cameras, i'd probably really enjoy advertising.
My naive hope is that eventually, spam-style ads will go away due to market forces. People with legitimate products will understand that more effective ad techniques exist, and shit-peddlers will be marginalized, much luck the current crop of spammers have been.
Suppose, for instance, that there is a new display driver model in Vista, and that DX10 works only on this new model. Perhaps DX10 and the new model were developed together, since in Vista so much of the UI is using the GPU at all times.
XP would lack this driver model, making DX10 not run on XP.
If the above scenario were true, it would tend to suggest a technical reason, don't you agree?
Do you know for a fact that what I've described isn't true?
Everytime I've got one of the desktops I support running something that requires a dip into admin priveleges for the apps that can't run in the user space, the OS is going to ask for verification. Given this will be *very* annoying, I'm guessing there's a little checkbox to "remember" this decision. Lo and behold! The system is running in Admin!
So let me see if I understand this: You are discussing a "problem" with an approach that you are speculating might work a certain way, on a feature and operating system you haven't ever used?
There has been a lot of work to improve the admin problem in Vista, and there's probably more that you don't see than what you do. Please don't make up your mind on what the drawbacks of the approaches we've taken until you've at least tried - and maybe understand - them.
turn the PC into a DRM'd set-top box.
No rational person thinks this, but suppose anyway that that is our secret plan, and that we're going to come up with some scheme whereby apps can't run unless they're magically signed or some other scheme.
Guess what - we already have that, in a few forms even (i.e. SAFER, SRP, etc), and the majority of people don't use it, and don't want to, and even if we did have it, there will still need to be a box that says "run anyway". So "turning the PC into a DRM set-top box" doesn't even solve the problem you're suggesting exists (which, in reality, doesn't exist, fyi)
Well, I have no reason to beleive it's fixed, but that's irrelevant. I specified that a specific version that I had experience with exhibited a problem, just like many other versions of other platforms have the same problem. Your brilliant retort amounts to "it is ridiculous to mention that something has a problem when there is something newer that doesn't". I disagree - i mention the Vista shell copy improvements precisely because i've had a lot of bad luck with shell file copying in the past, and this has been a multi-platform frustration.
That you suggest that it _was_ broken means you're disagreeing with what some other Apple fans have said. Or was it fine back then, and now it's even better? Or what?
Regarding telling me whats wrong with "XP security" - I'm listening if you've got workable solutions. Infact, if you can solve all our security challenges without breaking important customer scenarios, and make us some money in the process, we'll pay you any amount of money you can dream up. Or do you have some security advice that we haven't heard before? If you have some specific XP security problem you're running into please let us know.
I don't actually enjoy getting in grudge matches with people, and I am not really sure what you disagree with me on. We both apparently think that the file copy stuff in 10.3 was subpar (since by your admission, it is now "fixed"). Why are you attacking me? Is it because I had the "audacity" to point out what I have experiened in OS X and beleive is a shortcoming?
That word you keep using.. I don't think it means what you think it means.
10.3 is what I have install media for. 10.3 is what came with this machine.. when it was purchased new about 2 years ago. 10.4 is barely a year old, afaik.
I don't know that 10.4 does this any better, I just have more experience with 10.3, so that's what I was describing.
Apple users do not need to have a persecution complex. I was merely stating my observation that more often than not, I am dissatisfied with shell file copies on large files, especially when networks are involved, and that OS X appears to be no exception when it comes to exhibiting this problem.
Obviously shell file copy performance can't be the only dimension of the merit of an operating system or else everybody wouldn't be so awful at it after so many years.
That was my approach in previous builds when the find either didn't work or i didn't realize that it worked:)
I find little honor or value in remembering where to look for things or how the internals of this stuff works in my day to day life. When I think like a software guy, it's handy/efficient to have remembered how it all works, but when i try and compartmentalize myself into a "normal user", i have to ask - why the hell should i learn all of these annoying details of the intriciacies of things that i don't even want to mess with in the first place?
IOW - i don't want to remember where i set sound schemes. Fixing settings like that is an interrupt driven task - i realize something is annoying me or needs to be changed, and then i want to get it unannoying in the least time possible, in a manner that causes me the least distraction, and that takes essentially no mental effort. I want to think about other things, not about how my computer works.
I just tried this again as I was typing this post. I opened control panel, hit classic view,then scrolled to the bottom and started looking through the icons starting with "s". Sound... Sound.. wtf is sound? I scrolled through all of the control panel icons looking for something I might have missed. Eventually, I gave up and typed "sound" in the search box. Oh, i see today that it is called "Audio Devices and Sounds" or something like that.
For me, the intuitive behavior (that sounds are set under something called "sounds") is actually not what happens. I don't want to reprogram my own brain to adapt to the computer, i want the computer to adapt to me.
Congratulations Microsoft! You have eventually caught up with OS/2 which had this back in 1992.
Tell me about it. BITD I downloaded the 2.11 CSD on my 14.4k modem to update my 2.1 install on my 8mb 486.
I loved OS/2 until I realized that I was never going to find decent modem / bbs software for it, and thus, all of my modem usage would be running under a virtual dos box. As you surely recall, DOS mode serial port usage _crushed_ OS/2.
Once I realized that there were no apps for OS/2 anyway, but that I didn't really miss them, and that virtualized apps sucked if they dealt with the serial port, I figured i'd look at Linux more seriously. Once I saw that linux was just like OS/2 - fast, nerdy, no apps - but that accessing the serial port with usable tools didn't simply crush the box - i was hooked.
IMO, MS did not ship any acceptable operating systems between DOS 5.0 and Windows 2000.
My aim here on slashdot is not to be a Microsoft fanboy - that's certainly not what I do at work. But - when we do something well (even well compared to how we used to do it, which may ultimately mean we're still doing it badly, but just less so;), I don't mind saying so. I have to use the same software everyone else does, and its only been in the last few years that I personally felt like it was finally good enough for me to bother with. I want our stuff to be better, so yeah, when I see us start to get something right that should have been working 14 years ago, as you point out, i am happy about it:)
I did this just last evening on my wife's iBook G4 800 that i was rebuilding (hard drive failed and apple refused to honor the warranty, newegg to the rescue..)
I mounted my windows XP box from the Mac and was copying.dmg files off of it over the network. The copy progress bar was hung for over 30 seconds at one point; i was trying to move/cancel it to no avail. Eventually it started painting again - 20MB of progress later. During this time the Finder window looking at the same SMB share also became non responsive.
Machine was a 1 day old install of OS X 10.3 with all software updates.
You may very well be correct - that file copying works great for you. And I am not saying that it's a universally bad experience for meither. But last nite it was. And other times when dealing with network mapped stores, I've had similar behavior, on NT, W2k, XP, and yes, OS X.
I partially agree with you, and because of my unix background, I am running vista as a non-priviledged user.
There are two aspects of this. The first is that, if you truly are running as a low-priv user, you need to get elevation prompts at the correct times to be able to live life. This works pretty well, although I keep a cmd.exe window running as local admin sitting around sometimes.
The other aspect of this, however, is that in the real world, a lot of people just dont run as admin, and a lot of apps just can't. So a bunch of work has gone into making admins "virtual admins", so to speak, where operations that actually require priviledge use still involve user interaction/confirmation.
In that sense, people running "as admin" are getting the customer experience - and internally, the way the "did you really want to do this, Mr. Admin?" stuff works is passionately debated:)
My opinion is that people are complaining about the wrong problem - as we continue to eliminate things that require priviledge use, the amount that we have to care about putting up with a just-in-time priviledge escalation model goes down.
I was just thinking "i know i like it better, but really, what do i like better about it?"
Then something occured to me.
Right now, i am copying 4GB of files off a usb disk to a network share. The shell file copy stuff has been completely re-worked (shell file operations has always been something that i have hated)
In vista, you get an expand/collapse pane to get details of what it is doing, and it seems to happen in its own thread. The copy dialog window shows up as its own window that you can minimize/restore/whatever, and best of all, it doesn't hang/slow down the shell in any way.
Note that XP and OS X (as of 10.3) get this badly wrong - the file copy dialog in both tends to be slow to repaint itself or to respond to window messages, and if you use a separate explorer/finder window to try and access the destination you're copying to, the window lurches slowly to try and redraw.
Not so with Vista.
So there you go - here is something that was so annoying to me in XP that I had just stopped using the shell to do any sort of large file operation - i'd break out cmd.exe and xcopy. Vista has fixed at least some of the file copy problems very admirably.
There are a lot of cool "small" things that I see, but maybe you have to be kind of nerdy to apprecate them? The task manager has some cool features on the build I am running. The eventviewer (eventvwr) is a completely new animal and is way cooler than the old one
A nice use of the pervasive desktop search integrated into the explorer windows is in Control Panel. We're pretty good about changing control panel wildly between releases, and I never remember which menu your system environment variables or enabling remote desktop or changing it so that the "Explorer:Start Navigation" sound is (none). Now i just hit "start->control panel", click in the search box for something like "sound" and i get search-as-i-type results that are pretty accurate and take me right to the control panel i want to go to.
Is any of that a big deal? No. Does it make me love Vista when i think about how much i hated doing that stuff on XP?
Yes
Apparently, there are a lot of "big" changes under the hood of Vista, but you don't always see them in a big way.
its lousy that he had such a hard time with it, but i can't say that his experience is representative.
I've been running Vista builds on my Dell D600 laptop for a couple months. I just insatlled a new build yesterday. Out of the box, my Intel 2100 802.11 card, my built-in Smartcard Reader, and my Conexant v90 modem were not detected. Everything else was, including my video hardware (Radeon Mobility). I didn't have any driver disks or anything, and it was a clean install.
The "get new drivers for my broken hardware" feature in vista works pretty slick. I used the Welcome center, clicked on "setup devices" (or something), and it went off to windows update and found and installed drivers for the 3 devices that weren't in-box.
Just for a point of comparison, if i install XP on the same exact machine, i have - vga video - no audio - no modem - no wireless net - no WIRED net - no SMBus controller - no smartcard reader
So i guess i am saying, not only does Vista work "not bad" on laptops (i've had some problems with sleep/resume that we're having trouble tracking down), but the out-of-box driver experience is BETTER than it is for XP. And my laptop is an old crusty one - my video subsystem doesn't support the new video subsystem or any of that fanciness.
My laptop is my main machine for all email, web surfing, office apps, code reviews, etc etc. And it runs vista 100% of the time (except when i boot into xp to setup for taking a newer vista build). I use it at home and at work, both on wireless networks, in very different configurations.
It is not without its irritations and rough edges, but it's good enough for me to use it for a significant portion of my work duties.
I am curious about what build this guy was running - I'm surprised that he's been working on it for "4 days" because i'm not sure the final Beta2 build has been public that long:)
I see that you enjoy being pejorative towards ND. On behalf of the residents of Fargo, we're cautiously optimistic that you're planning on staying in Boston. People that can tolerate or even thrive in New England (especially the Peoples Republic of Massechussets -- there's me returning the favor) probably can't tolerate Fargo, so I'd guess that you and I have made the correct decisions about where each of us should live, respectively:)
Fwiw, the Fargo/Moorhead/WestFargo "metro area" has something like 180k people. It's not exactly rural - especially compared to the rest of the state.
The amusing thing is that for many residents, Fargo "is" the big city - there are only a handful of towns within a days drive that have a 5 digit or higher population. People move to Fargo from outlying areas and some of them can't handle the size.
My wife and I moved here from Seattle. The thing I miss most is the lack of insects. Our house didn't have screens on the windows because it didn't need them. The mosquitos here are quite unpleasant...it really dampens the ability to enjoy the nice seasons (you get all 4 of them here, also unlike Seattle)
The nice thing about the US is that there's a huge variety of people, and there seems to be a place that's tailor made for just about everybody's natural comfort level. This might sound like segregation or several other words with negative connotations, but I'd claim it's more like "Mosaic Of Subcultures", but at a macroscopic level. (See Christopher Alexander if you don't catch the reference)
When taken aggregately, state wide, it certainly is one of the coldest places. We regularly stay in the -30F range for a couple days at a time, and then we'll have 30mph winds that go with it. Fyi, exposed skin in those conditions isn't good for more than a couple of minutes until its frostbitten. When I walk from my car in the parking lot to my building entrance, wearing a full face mask with eyeholes, i have a headache by the time i get to the door, and the bridge of my nose is numb.
There are days that the NWS says "stay inside today".
This doesn't begin to involve snowy conditions. All the interstate on ramps have manual train-crossing-style gates across them. The highway patrol will block entrance to the interstates if the right combinations of wind, cold, and snowcover converge. That's because if there's an accident, nobody will be able to come get you.
Someone else was talking about some kind of freak mountain that has a lot of bad weather and accidents. Ok, sure, and Mt Everest is also cold and dangerous. It's a mountain, what do you expect? People dont do mountain climbing because its low-impact cardio.
North Dakota is one of the only places I'm aware of in the 48 contiguous states where there's nothing to hit, no elevation changes, no hurricaine/volcano/earthquake issues, and yet still has a high risk of weather-related death. This is not about looking for adventure on some mountain - there are days when you decide that taking the flat, straight, well paved highway is too high of a risk just to get to the grocery store.
Is not the legislative branch of the US government the body that is supposed to be responsible for passing laws to protect our freedom and liberty?
No. The constitution and our armed forces protect our freedom and liberty. The legislative branch of the US government would serve us best by not doing _anything_, since by definition each new law they add restricts somebodies freedoms. The only pass here is when they make a law that restricts one freedom because it is ultimately denying someone else a more important freedom (you do not have the freedom to murder me because you would deny me my protected rights to life, leisure, and persuit of happiness)
The founding fathers knew that the legislature was not to be trusted, and therefore gave the judiciary oversight on every law the legislative branch came up with, allegedly to ensure that none of our constitutionally protected rights are ever trampled.
What infact happens is that the legislature creates laws to please their campaign contributors, the judicial branch varies between de-facto law writing and waving their hands and saying "we cant do what's proper because this law is actually as stupid as you think it is, but it's the law".
Let's not mention the executive branch.:)
On the matter of the specific issue -
the market would sort this entire issue out completely if the entities in question were not government granted monopolies. How does AT&T own the last mile? Government easements that let them bury cable where they want it. How does Verizon lock Vonage out of the market? Pushing for "tariff equality" and 911 service - both government inventions, not market ones. Why do most people have a consumer-only internet (by this i mean - poor upstream speed, all interesting ports blocked, etc)? Because IP service is quickly consolidating due to government barriers to entry (running cables, using wireless, etc etc) Why does Verizon customer service suck, why do they only sell crippled phones, and why are they focused on extortion-style lock-in contracts? Because the government created barriers to entry for running a cell provider are higher than they need to be.
Companies are immoral entities that exist to make as much money as possible. Manipulating the marketplace via government intervention is one of the best ways they acheive that - the government is the only factor that trumps the market, the only place where the customer has _zero_ choice.
As I understand it, recoding Immo2 and Immo3 systems with VAG COM requires an SKC, which only a dealership with a terminal to the VAG mothership in germany, can give you. The latest generation SKC is also datestamp dependant, so it's only valid on the day it was requested.
The dealer is under no obligation to give you the SKC, but, assuming they do, THEN you can recode to your hearts conteng using VAS-1551, VAG-COM, or whatever.
I think the insinuation of the OP was that engine complexity has risen sufficiently that you will not get the car started without ECU cooperation. I tend to agree.
The line to draw here might be in the mid-late 80s, when prior to that, some cars had mechanical primary operation with electronic "tuning", but a non-ECU limp-home mode. Bosch KE-Jetronic (CIS-E) comes to mind. Even newer ignition modules have a limp-mode where timing is a static 10BTDC.. ala Ford EDIS.. it will still send a spark signal even in the absense of the advance setting from the ECU.
But even as far back as Motronic 1.0 (mid 80s), if the ECU didn't feel like sending a spark signal, there wasn't going to be one. THere was no anti-theft in M1.0, but the point remains - the ECU has final say in the startability of any sufficiently modern car. (note its worth making the distinction between an ECU that withholds a start signal and one that also withholds ignition/fuel signals on a cranking engine.. an effectivee attack on some cars is to manually engage the starter.. the ECU sees the engine "cranking" and supplies the necessariy signals to the injectors and ignitor(s))
To work around an uncooperative ECU you need to come up with your own fuel control at a minimum (maybe you could fake this by jumpering the fuel pump relay, and then having a potentiometer driving an oscillator on the injectors, assuming batch injection) and spark, assuming the built in ignition doesn't have a fall-back mode.
I think we agree that the ECU being uncooperative makes life worse on old and new cars alike, but i think that there is a sufficiently old/dumb system out there that the ECU can be worked around, where as a sufficiently modern system, the ECU must be convinced to cooperate, as a practical matter.
I'll respond to this because it's interesting. Point-taken about getting over myself, although it would help if you could be more specific. What should I have not done or said? Are you just badgering me for the fun of it?
And yet it hasn't stopped you, here
I am trying to figure out: in a post that largely recounts my personal experiences working at MSFT, how i am talking about stuff i don't understand/know about?
Your degree in is what, exactly?
Thanks for asking. I took the easy route and got a double major in Math and Computer Science. I was admitted to state university (Nebraska. Hold laughter, please;) on a full academic scholarship into their engineering program. I had 20 hours - including 3 semesters of calculus -- completed due to AP classes and transfer credits when I showed up on day one.
After my first year of circuits classes, I decided that the results I was getting in those classes were not justifiable given the amount of time I was dedicating to them. (IOW: I found them more challenging than I cared to deal with. Most people in this engineering program were on a 5 year plan; I wanted to be done in 4 and out in the work force full time).
So, I dropped out of the engineering college and went for the math/compsci double major as an easy-out (since i had so much math done by the time I entered school)
You're a god damn tester
Yep. And at the places I worked before MS (private consulting, Boeing, a national ISP, university IS department), I was either a developer or a UNIX admin, or usually both. At one company, I rewrote their C++ billing system (that they had contracted out and paid a quarter-million dollars foar) and made it run in a fraction of the time and improved the accuracy, and expanded its features so that the ISP could offer more types of products and plans. In 3 months, because I was still a college student during the school year.
It was only when I got to Microsoft that I met developers that said "you're not good enough to be one of us" (not worded quite that way, mind you).
So, I guess maybe the claim here is that people that are good enough to be star developers at a lot of places aren't qualified to be developers at all at Microsoft. Although as I wrote elsewhere, I still end up writing a lot of code, that gets checked into the same depot as dev code, and has to meet the same quality bar.
For a while I was kind of hung up on the dev vs sde/t thing. People have moved in both directions between the two (which is considered a lateral move, btw, even in payscale). Originally I was thinking I'd try to move towards a developer type role, but I am over the idea. I care more about making sure that MS delivers products that infuriate customers as little as possible.. as opposed to worrying about what an A/C on slashdot has to say about my job title.
I don't have a really good answer for this, but you can try asking them the following litmus questions.
1) Are you sure? Why? 2) Has that been your personal experience? No? Where did you hear about it? Why do you beleive it enough to be retelling it? 3) Is this something that's even measurable? How did you measure it? Why do you think that is the appropriate way of measuring it?
So the example I gave earlier on.. I was telling this NT guy that making GDI or Win32 (or something -- i dont even remember) in-kernel for NT4 was a dumb move. Byte magazine said so, they showed the pretty rectangles on the architecture diagram. Clearly, this was introducing instability into the system, because if Win32 crashed it would bring down the system! Horrors!
That's what I read, that's what I repeated, that's what this interviewer just tolerated. It's too bad he didn't bust my chops right then and there.
Years later, as I was reading Inside Windows 2000 3rd Ed, written by somebody that actually knows wtf they're talking about (and has an NT source license, and is the guy behind sysinternals.com), he was talking about this very problem. And how in every version of NT, if CSRSS (Win32 subsystem runtime) crashed/returned to SMSS.exe (Subsystem/Session Manager - responsible for starting the Win32, OS2, and Posix subsystems in NT), SMSS would bug check the box immediately.
Boy, that's an interesting detail. So it didn't matter where they put Win32 stuff - if there was a crash, the box was going down automatically, by design. The Win32 -> kernel move was a big perf win (apparently), and NT has only gotten more and more stable over time, not less so.
(for those interested, this explanation appears in the grey box starting on pg 53 of the book)
When i think about this story, any of those litmus test questions would have caught it. Did i really understand the problem well enough to be talking about it? No. Had Win32 crashing been my personal experience? No - I'd refused to even use NT up until that point in my life. Why was I even bringing this up? To sound intelligent. To try and be provocative. Because Microsoft was "the enemy".
Now, when I talk about this story, I make sure to attribute the "knowledge" to the source I got it from (The "Book"). Infact, I even pulled it out while writing this to make sure I got the details acceptably close. I still don't have any personal experience with CSRSS crashing. But I "buy" the argument presented in the book until I learn otherwise (by hearing a better argument).
Slashdot Mirror
The problem is that the "maximum you are willing to pay" assumption is flawed.
I don't know the maximum I am willing to pay, because I don't know the "value" of an item. Why don't I know the value of the item? Because I don't know what other people are willing to pay for it. What the marketplace is willing to pay for something determines its value. I can't choose a bid ahead of time that will guarantee that I win the auctino yet still guarantee that I pay something I can afford.
If my max bid is in the ballpark of what I'll spend, I'll still spend 1% extra to outbid someone else, thus blowing my "maximum".
I chould choose, $20, $80, or $200 as a max bid for something. But at any of those price points, I'd be willing to pay $21, $81, or $201, because it represents such a trivial incremental addition to the price. If i want the thing bad enough to bid on it, I want to win it. Quibbling over a few dollars is less important than actually winning.
After I lost a few Cisco 678 auctions, I just sniped the next one that came up (and paid less, but that was not my motivation for doing it)
I am a big fan of the _idea_ of public transit.. I spent a few weeks in Munich and it was wonderful..but munich has, iirc, 9 u-bahn and 27 s-bahn lines for a metro area of about 1 million people. Isn't Marienplatz or Munich Hauptbanhof 4-5 levels deep ? I have no idea how a population of 1m people can support such an incredible public train system (but I sure enjoyed it while I was there!).
When we went to Berlin (via car - we toured Germany via car and I also had a quick stop to drive 6 laps of the Nordschleife) we just parked at the most distant P&R we could fine. Touring any large German city via car is simply pointless. The U-bahn system in berlin was very poor compared to Munich, but then, they're still re-assembling things in light of the 1989 re-unification.. and their system is much older and has more legacy-inspired problems afaict.
Now, given that Germany has excellent mass transit in dense cities... but also the most excellent highway system on the planet.. doesn't it seem like good public transit in beautiful dense cities and excellent highway systems are not mutually exclusive? Fwiw, Germany also has the ICE rail system which is frankly faster than any car you can rent easily.
Now, onto the US.
Eisenhower's goal was military, with the civilian benefits being just that - benefits. Next time you're scooting along the interstate in rural america.. going 70-80mph, imagine trying to go even 1/3rd that speed on the ground even 10feet away from the edge of the road.. it doesn't matter what kind of vehicle we're talking about; moving over unimproved ground at any kind of speed is near impossible, and certainly hard on equipment and people.
Christopher Alexander points out many of the shortcomings you do w.r.t. our car-focused society. My wife and I live "downtown" and we have to drive out to the subburbs/strip malls to buy groceries (although last evening we bicycled). It was after our Germany trip that we realized how good it feels to walk places instead of just hopping in the car. But you (and Alexander) point out many of the defects in American city/road planning that cause them to be car-optimized and pedestrian antagonistic.
It's hard to say that Ike made a "mistake". It's not a forgone conclusion that the US would have developed identically to europe even without our interstate system. And it's also not a forgone conclusion that the US would not have suffered some of the same ills without an interstate system. For instance, though the smog in LA is bad, isn't the air in London pretty bad also? And don't they have a pretty extensive underground rail system? Presumably, fewer Londoners depend on personal automobiles for daily commuting yet the air is still (reputedly) poor.
FAT32 and NTFS, are ages behind file systems like ResierFS (especially reiserfs v4) and even Ext3
Beleive me, I'm no huge fan of NTFS, but could you explain why Ext3 is ages ahead of it?
There's a lot of interesting work going on Reiser - that I'll certainly agree with. Although if anecdotes about people losing data with it are any indication, "stability" isn't one of the advantages it has over NTFS (at least for now..)
The only people that have been in the US for more than about 30 generations are laregly confined to "reservations", which are hotbeds of crime, inebriation, and a dead-end life.
(at least, that's my impression from visiting a few reservations, including the tribe of which my great grandmother was 100% full blooded and where I probably have a few distant relatives...)
The idea that you are "more american" because your family has been here for "several generations" is dubious. How many generations must have been born here for you to be a "real American" ?
In my opinion, anyone that is willing to follow the laws of this country should be able to become a citizen, easily and cheaply. If you think your birth pedigree is of relevance in what your rights should be, you are hardly living in the spirt of the principles on which America was founded.. perhaps you'd be happier in colonial/monarchial Europe of a few hundred years ago, or India under the caste system?
The immigration policy of this country should be guided by the following principle: "If you are willing to come here legally and comply with our laws, stay as long as you like"
There is no search engine policy at Microsoft :)
:)
That said, depending on who you are talking to, some employees will say "use the MS product in space blah no mater what.. because if you don't think it's the best for you, you need to help us understand how to make it better"
Other employees say "sometimes I don't want to be Microsoft's First Customer - I just want to do my thing without any hassles"
Very few employees are 100% MSFT all the time for all topics. Google _is_ a popular search destination. The issue of Google vs MSN/windows live search is a sore one for lots of employees. The guys behind the search technologies have setup some clever multi-search stuff internally where you can vote on results from the various different engines (which does account for some fraction of any microsoft-hitting-google traffic, btw) and then they decide how to roll human selectivity feedback into the technology. They'll tell everyone to use our stuff, even if its via one of the multi-search proxies, because they want all the data/criticism they can get to try and make it work better. The employees that just use google aren't thrilled either - tmore often than not, hey'd _like_ to use our stuff, but they just dont have the patience to try something else first, then get frustrated with the results, THEN go to Google.
In my years at Microsoft, I have never seen any evidence to indicate that we have any kind of outbound filteirng whatsoever. Certainly there is proxying and logging, but I've never seen anything blocked. I think the policy must be something like "we assume our employees are reasonable, intelligent people. We can give them free reign and if they abuse that priviledge we have logs to support any disciplinary action we may take"
Also, for essentially any content you can think of, it is somebody's job to look at that content. (eg the IE test team needs to make sure google renders properly...maybe people in our MSN properties need to be able to look at "inappropriate" material at times as part of cooperation with FBI investigations.. our developers are not allowed to look at online patent docs or GPL source code (for fear of IP pollution lawsuits hitting us), but our legal team HAS to in order to do some of thier job...)
Figuring out what should be filtered for what people would be a lot of effort. It's probably easier to just fire people that do things they shouldn't
I was thinking of posting something like this. Something like "I am sure the Chinese spec has monitoring backdoors in it known only to the Chinese govt". I'd say this because I'd be implying that whatever the US is in favor of wouldn't have such drawbacks.
Except I don't think it's reasonable to think or say such a thing in light of recent events in the US, which is a shame.
I was going to ask the same question. A friend of mine works on adserving technology and i was giving him a bit of greif about how he slept at nite, and so on.
He and I are both car junkies so he had a clever response. "If, when you saw ads, they were things like new products for your specific car, would you be as mad at them? I mean, if someone makes a new 9lb flywheel for your engine, and we show you that ad, will it be upsetting?"
I had to concede - no. I currently spend my time trying to find what I want when it comes to go-fast parts for my cars.
If I only ever saw ads for performance car parts for cars that I own, deals on new anime releases, and accessories for canon EOS cameras, i'd probably really enjoy advertising.
My naive hope is that eventually, spam-style ads will go away due to market forces. People with legitimate products will understand that more effective ad techniques exist, and shit-peddlers will be marginalized, much luck the current crop of spammers have been.
How do you know?
Suppose, for instance, that there is a new display driver model in Vista, and that DX10 works only on this new model. Perhaps DX10 and the new model were developed together, since in Vista so much of the UI is using the GPU at all times.
XP would lack this driver model, making DX10 not run on XP.
If the above scenario were true, it would tend to suggest a technical reason, don't you agree?
Do you know for a fact that what I've described isn't true?
Everytime I've got one of the desktops I support running something that requires a dip into admin priveleges for the apps that can't run in the user space, the OS is going to ask for verification.
Given this will be *very* annoying, I'm guessing there's a little checkbox to "remember" this decision. Lo and behold! The system is running in Admin!
So let me see if I understand this: You are discussing a "problem" with an approach that you are speculating might work a certain way, on a feature and operating system you haven't ever used?
There has been a lot of work to improve the admin problem in Vista, and there's probably more that you don't see than what you do. Please don't make up your mind on what the drawbacks of the approaches we've taken until you've at least tried - and maybe understand - them.
turn the PC into a DRM'd set-top box.
No rational person thinks this, but suppose anyway that that is our secret plan, and that we're going to come up with some scheme whereby apps can't run unless they're magically signed or some other scheme.
Guess what - we already have that, in a few forms even (i.e. SAFER, SRP, etc), and the majority of people don't use it, and don't want to, and even if we did have it, there will still need to be a box that says "run anyway". So "turning the PC into a DRM set-top box" doesn't even solve the problem you're suggesting exists (which, in reality, doesn't exist, fyi)
Well, I have no reason to beleive it's fixed, but that's irrelevant. I specified that a specific version that I had experience with exhibited a problem, just like many other versions of other platforms have the same problem. Your brilliant retort amounts to "it is ridiculous to mention that something has a problem when there is something newer that doesn't". I disagree - i mention the Vista shell copy improvements precisely because i've had a lot of bad luck with shell file copying in the past, and this has been a multi-platform frustration.
That you suggest that it _was_ broken means you're disagreeing with what some other Apple fans have said. Or was it fine back then, and now it's even better? Or what?
Regarding telling me whats wrong with "XP security" - I'm listening if you've got workable solutions. Infact, if you can solve all our security challenges without breaking important customer scenarios, and make us some money in the process, we'll pay you any amount of money you can dream up. Or do you have some security advice that we haven't heard before? If you have some specific XP security problem you're running into please let us know.
I don't actually enjoy getting in grudge matches with people, and I am not really sure what you disagree with me on. We both apparently think that the file copy stuff in 10.3 was subpar (since by your admission, it is now "fixed"). Why are you attacking me? Is it because I had the "audacity" to point out what I have experiened in OS X and beleive is a shortcoming?
That word you keep using.. I don't think it means what you think it means.
10.3 is what I have install media for. 10.3 is what came with this machine.. when it was purchased new about 2 years ago. 10.4 is barely a year old, afaik.
I don't know that 10.4 does this any better, I just have more experience with 10.3, so that's what I was describing.
Apple users do not need to have a persecution complex. I was merely stating my observation that more often than not, I am dissatisfied with shell file copies on large files, especially when networks are involved, and that OS X appears to be no exception when it comes to exhibiting this problem.
Obviously shell file copy performance can't be the only dimension of the merit of an operating system or else everybody wouldn't be so awful at it after so many years.
That was my approach in previous builds when the find either didn't work or i didn't realize that it worked :)
,then scrolled to the bottom and started looking through the icons starting with "s". Sound... Sound.. wtf is sound? I scrolled through all of the control panel icons looking for something I might have missed. Eventually, I gave up and typed "sound" in the search box. Oh, i see today that it is called "Audio Devices and Sounds" or something like that.
I find little honor or value in remembering where to look for things or how the internals of this stuff works in my day to day life. When I think like a software guy, it's handy/efficient to have remembered how it all works, but when i try and compartmentalize myself into a "normal user", i have to ask - why the hell should i learn all of these annoying details of the intriciacies of things that i don't even want to mess with in the first place?
IOW - i don't want to remember where i set sound schemes. Fixing settings like that is an interrupt driven task - i realize something is annoying me or needs to be changed, and then i want to get it unannoying in the least time possible, in a manner that causes me the least distraction, and that takes essentially no mental effort. I want to think about other things, not about how my computer works.
I just tried this again as I was typing this post. I opened control panel, hit classic view
For me, the intuitive behavior (that sounds are set under something called "sounds") is actually not what happens. I don't want to reprogram my own brain to adapt to the computer, i want the computer to adapt to me.
Congratulations Microsoft! You have eventually caught up with OS/2 which had this back in 1992.
;), I don't mind saying so. I have to use the same software everyone else does, and its only been in the last few years that I personally felt like it was finally good enough for me to bother with. I want our stuff to be better, so yeah, when I see us start to get something right that should have been working 14 years ago, as you point out, i am happy about it :)
Tell me about it. BITD I downloaded the 2.11 CSD on my 14.4k modem to update my 2.1 install on my 8mb 486.
I loved OS/2 until I realized that I was never going to find decent modem / bbs software for it, and thus, all of my modem usage would be running under a virtual dos box. As you surely recall, DOS mode serial port usage _crushed_ OS/2.
Once I realized that there were no apps for OS/2 anyway, but that I didn't really miss them, and that virtualized apps sucked if they dealt with the serial port, I figured i'd look at Linux more seriously. Once I saw that linux was just like OS/2 - fast, nerdy, no apps - but that accessing the serial port with usable tools didn't simply crush the box - i was hooked.
IMO, MS did not ship any acceptable operating systems between DOS 5.0 and Windows 2000.
My aim here on slashdot is not to be a Microsoft fanboy - that's certainly not what I do at work. But - when we do something well (even well compared to how we used to do it, which may ultimately mean we're still doing it badly, but just less so
Fabulous work, Dr. Holmes.
:)
The "we" was by no means a slipup - look at my slashdot user info. You didn't need to bother looking at my website
I did this just last evening on my wife's iBook G4 800 that i was rebuilding (hard drive failed and apple refused to honor the warranty, newegg to the rescue..)
.dmg files off of it over the network. The copy progress bar was hung for over 30 seconds at one point; i was trying to move/cancel it to no avail. Eventually it started painting again - 20MB of progress later. During this time the Finder window looking at the same SMB share also became non responsive.
I mounted my windows XP box from the Mac and was copying
Machine was a 1 day old install of OS X 10.3 with all software updates.
You may very well be correct - that file copying works great for you. And I am not saying that it's a universally bad experience for meither. But last nite it was. And other times when dealing with network mapped stores, I've had similar behavior, on NT, W2k, XP, and yes, OS X.
I partially agree with you, and because of my unix background, I am running vista as a non-priviledged user.
:)
There are two aspects of this. The first is that, if you truly are running as a low-priv user, you need to get elevation prompts at the correct times to be able to live life. This works pretty well, although I keep a cmd.exe window running as local admin sitting around sometimes.
The other aspect of this, however, is that in the real world, a lot of people just dont run as admin, and a lot of apps just can't. So a bunch of work has gone into making admins "virtual admins", so to speak, where operations that actually require priviledge use still involve user interaction/confirmation.
In that sense, people running "as admin" are getting the customer experience - and internally, the way the "did you really want to do this, Mr. Admin?" stuff works is passionately debated
My opinion is that people are complaining about the wrong problem - as we continue to eliminate things that require priviledge use, the amount that we have to care about putting up with a just-in-time priviledge escalation model goes down.
I was just thinking "i know i like it better, but really, what do i like better about it?"
Then something occured to me.
Right now, i am copying 4GB of files off a usb disk to a network share. The shell file copy stuff has been completely re-worked (shell file operations has always been something that i have hated)
In vista, you get an expand/collapse pane to get details of what it is doing, and it seems to happen in its own thread. The copy dialog window shows up as its own window that you can minimize/restore/whatever, and best of all, it doesn't hang/slow down the shell in any way.
Note that XP and OS X (as of 10.3) get this badly wrong - the file copy dialog in both tends to be slow to repaint itself or to respond to window messages, and if you use a separate explorer/finder window to try and access the destination you're copying to, the window lurches slowly to try and redraw.
Not so with Vista.
So there you go - here is something that was so annoying to me in XP that I had just stopped using the shell to do any sort of large file operation - i'd break out cmd.exe and xcopy. Vista has fixed at least some of the file copy problems very admirably.
There are a lot of cool "small" things that I see, but maybe you have to be kind of nerdy to apprecate them? The task manager has some cool features on the build I am running. The eventviewer (eventvwr) is a completely new animal and is way cooler than the old one
A nice use of the pervasive desktop search integrated into the explorer windows is in Control Panel. We're pretty good about changing control panel wildly between releases, and I never remember which menu your system environment variables or enabling remote desktop or changing it so that the "Explorer:Start Navigation" sound is (none). Now i just hit "start->control panel", click in the search box for something like "sound" and i get search-as-i-type results that are pretty accurate and take me right to the control panel i want to go to.
Is any of that a big deal? No. Does it make me love Vista when i think about how much i hated doing that stuff on XP?
Yes
Apparently, there are a lot of "big" changes under the hood of Vista, but you don't always see them in a big way.
its lousy that he had such a hard time with it, but i can't say that his experience is representative.
:)
I've been running Vista builds on my Dell D600 laptop for a couple months. I just insatlled a new build yesterday. Out of the box, my Intel 2100 802.11 card, my built-in Smartcard Reader, and my Conexant v90 modem were not detected. Everything else was, including my video hardware (Radeon Mobility). I didn't have any driver disks or anything, and it was a clean install.
The "get new drivers for my broken hardware" feature in vista works pretty slick. I used the Welcome center, clicked on "setup devices" (or something), and it went off to windows update and found and installed drivers for the 3 devices that weren't in-box.
Just for a point of comparison, if i install XP on the same exact machine, i have
- vga video
- no audio
- no modem
- no wireless net
- no WIRED net
- no SMBus controller
- no smartcard reader
So i guess i am saying, not only does Vista work "not bad" on laptops (i've had some problems with sleep/resume that we're having trouble tracking down), but the out-of-box driver experience is BETTER than it is for XP. And my laptop is an old crusty one - my video subsystem doesn't support the new video subsystem or any of that fanciness.
My laptop is my main machine for all email, web surfing, office apps, code reviews, etc etc. And it runs vista 100% of the time (except when i boot into xp to setup for taking a newer vista build). I use it at home and at work, both on wireless networks, in very different configurations.
It is not without its irritations and rough edges, but it's good enough for me to use it for a significant portion of my work duties.
I am curious about what build this guy was running - I'm surprised that he's been working on it for "4 days" because i'm not sure the final Beta2 build has been public that long
I see that you enjoy being pejorative towards ND. On behalf of the residents of Fargo, we're cautiously optimistic that you're planning on staying in Boston. People that can tolerate or even thrive in New England (especially the Peoples Republic of Massechussets -- there's me returning the favor) probably can't tolerate Fargo, so I'd guess that you and I have made the correct decisions about where each of us should live, respectively :)
Fwiw, the Fargo/Moorhead/WestFargo "metro area" has something like 180k people. It's not exactly rural - especially compared to the rest of the state.
The amusing thing is that for many residents, Fargo "is" the big city - there are only a handful of towns within a days drive that have a 5 digit or higher population. People move to Fargo from outlying areas and some of them can't handle the size.
My wife and I moved here from Seattle. The thing I miss most is the lack of insects. Our house didn't have screens on the windows because it didn't need them. The mosquitos here are quite unpleasant...it really dampens the ability to enjoy the nice seasons (you get all 4 of them here, also unlike Seattle)
The nice thing about the US is that there's a huge variety of people, and there seems to be a place that's tailor made for just about everybody's natural comfort level. This might sound like segregation or several other words with negative connotations, but I'd claim it's more like "Mosaic Of Subcultures", but at a macroscopic level. (See Christopher Alexander if you don't catch the reference)
When taken aggregately, state wide, it certainly is one of the coldest places. We regularly stay in the -30F range for a couple days at a time, and then we'll have 30mph winds that go with it. Fyi, exposed skin in those conditions isn't good for more than a couple of minutes until its frostbitten. When I walk from my car in the parking lot to my building entrance, wearing a full face mask with eyeholes, i have a headache by the time i get to the door, and the bridge of my nose is numb.
There are days that the NWS says "stay inside today".
This doesn't begin to involve snowy conditions. All the interstate on ramps have manual train-crossing-style gates across them. The highway patrol will block entrance to the interstates if the right combinations of wind, cold, and snowcover converge. That's because if there's an accident, nobody will be able to come get you.
Someone else was talking about some kind of freak mountain that has a lot of bad weather and accidents. Ok, sure, and Mt Everest is also cold and dangerous. It's a mountain, what do you expect? People dont do mountain climbing because its low-impact cardio.
North Dakota is one of the only places I'm aware of in the 48 contiguous states where there's nothing to hit, no elevation changes, no hurricaine/volcano/earthquake issues, and yet still has a high risk of weather-related death. This is not about looking for adventure on some mountain - there are days when you decide that taking the flat, straight, well paved highway is too high of a risk just to get to the grocery store.
Is not the legislative branch of the US government the body that is supposed to be responsible for passing laws to protect our freedom and liberty?
No. The constitution and our armed forces protect our freedom and liberty. The legislative branch of the US government would serve us best by not doing _anything_, since by definition each new law they add restricts somebodies freedoms. The only pass here is when they make a law that restricts one freedom because it is ultimately denying someone else a more important freedom (you do not have the freedom to murder me because you would deny me my protected rights to life, leisure, and persuit of happiness)
The founding fathers knew that the legislature was not to be trusted, and therefore gave the judiciary oversight on every law the legislative branch came up with, allegedly to ensure that none of our constitutionally protected rights are ever trampled.
What infact happens is that the legislature creates laws to please their campaign contributors, the judicial branch varies between de-facto law writing and waving their hands and saying "we cant do what's proper because this law is actually as stupid as you think it is, but it's the law".
Let's not mention the executive branch.
On the matter of the specific issue -
the market would sort this entire issue out completely if the entities in question were not government granted monopolies. How does AT&T own the last mile? Government easements that let them bury cable where they want it. How does Verizon lock Vonage out of the market? Pushing for "tariff equality" and 911 service - both government inventions, not market ones. Why do most people have a consumer-only internet (by this i mean - poor upstream speed, all interesting ports blocked, etc)? Because IP service is quickly consolidating due to government barriers to entry (running cables, using wireless, etc etc) Why does Verizon customer service suck, why do they only sell crippled phones, and why are they focused on extortion-style lock-in contracts? Because the government created barriers to entry for running a cell provider are higher than they need to be.
Companies are immoral entities that exist to make as much money as possible. Manipulating the marketplace via government intervention is one of the best ways they acheive that - the government is the only factor that trumps the market, the only place where the customer has _zero_ choice.
As I understand it, recoding Immo2 and Immo3 systems with VAG COM requires an SKC, which only a dealership with a terminal to the VAG mothership in germany, can give you. The latest generation SKC is also datestamp dependant, so it's only valid on the day it was requested.
The dealer is under no obligation to give you the SKC, but, assuming they do, THEN you can recode to your hearts conteng using VAS-1551, VAG-COM, or whatever.
I think the insinuation of the OP was that engine complexity has risen sufficiently that you will not get the car started without ECU cooperation. I tend to agree.
The line to draw here might be in the mid-late 80s, when prior to that, some cars had mechanical primary operation with electronic "tuning", but a non-ECU limp-home mode. Bosch KE-Jetronic (CIS-E) comes to mind. Even newer ignition modules have a limp-mode where timing is a static 10BTDC.. ala Ford EDIS.. it will still send a spark signal even in the absense of the advance setting from the ECU.
But even as far back as Motronic 1.0 (mid 80s), if the ECU didn't feel like sending a spark signal, there wasn't going to be one. THere was no anti-theft in M1.0, but the point remains - the ECU has final say in the startability of any sufficiently modern car. (note its worth making the distinction between an ECU that withholds a start signal and one that also withholds ignition/fuel signals on a cranking engine.. an effectivee attack on some cars is to manually engage the starter.. the ECU sees the engine "cranking" and supplies the necessariy signals to the injectors and ignitor(s))
To work around an uncooperative ECU you need to come up with your own fuel control at a minimum (maybe you could fake this by jumpering the fuel pump relay, and then having a potentiometer driving an oscillator on the injectors, assuming batch injection) and spark, assuming the built in ignition doesn't have a fall-back mode.
I think we agree that the ECU being uncooperative makes life worse on old and new cars alike, but i think that there is a sufficiently old/dumb system out there that the ECU can be worked around, where as a sufficiently modern system, the ECU must be convinced to cooperate, as a practical matter.
You can download Virtual Server from Microsoft today, for free. I've been using it for over a year.
Nice work, anonymous antagonizer :)
;) on a full academic scholarship into their engineering program. I had 20 hours - including 3 semesters of calculus -- completed due to AP classes and transfer credits when I showed up on day one.
I'll respond to this because it's interesting. Point-taken about getting over myself, although it would help if you could be more specific. What should I have not done or said? Are you just badgering me for the fun of it?
And yet it hasn't stopped you, here
I am trying to figure out: in a post that largely recounts my personal experiences working at MSFT, how i am talking about stuff i don't understand/know about?
Your degree in is what, exactly?
Thanks for asking. I took the easy route and got a double major in Math and Computer Science. I was admitted to state university (Nebraska. Hold laughter, please
After my first year of circuits classes, I decided that the results I was getting in those classes were not justifiable given the amount of time I was dedicating to them. (IOW: I found them more challenging than I cared to deal with. Most people in this engineering program were on a 5 year plan; I wanted to be done in 4 and out in the work force full time).
So, I dropped out of the engineering college and went for the math/compsci double major as an easy-out (since i had so much math done by the time I entered school)
You're a god damn tester
Yep. And at the places I worked before MS (private consulting, Boeing, a national ISP, university IS department), I was either a developer or a UNIX admin, or usually both. At one company, I rewrote their C++ billing system (that they had contracted out and paid a quarter-million dollars foar) and made it run in a fraction of the time and improved the accuracy, and expanded its features so that the ISP could offer more types of products and plans. In 3 months, because I was still a college student during the school year.
It was only when I got to Microsoft that I met developers that said "you're not good enough to be one of us" (not worded quite that way, mind you).
So, I guess maybe the claim here is that people that are good enough to be star developers at a lot of places aren't qualified to be developers at all at Microsoft. Although as I wrote elsewhere, I still end up writing a lot of code, that gets checked into the same depot as dev code, and has to meet the same quality bar.
For a while I was kind of hung up on the dev vs sde/t thing. People have moved in both directions between the two (which is considered a lateral move, btw, even in payscale). Originally I was thinking I'd try to move towards a developer type role, but I am over the idea.
I care more about making sure that MS delivers products that infuriate customers as little as possible.. as opposed to worrying about what an A/C on slashdot has to say about my job title.
I don't have a really good answer for this, but you can try asking them the following litmus questions.
1) Are you sure? Why?
2) Has that been your personal experience? No? Where did you hear about it? Why do you beleive it enough to be retelling it?
3) Is this something that's even measurable? How did you measure it? Why do you think that is the appropriate way of measuring it?
So the example I gave earlier on.. I was telling this NT guy that making GDI or Win32 (or something -- i dont even remember) in-kernel for NT4 was a dumb move. Byte magazine said so, they showed the pretty rectangles on the architecture diagram. Clearly, this was introducing instability into the system, because if Win32 crashed it would bring down the system! Horrors!
That's what I read, that's what I repeated, that's what this interviewer just tolerated. It's too bad he didn't bust my chops right then and there.
Years later, as I was reading Inside Windows 2000 3rd Ed, written by somebody that actually knows wtf they're talking about (and has an NT source license, and is the guy behind sysinternals.com), he was talking about this very problem. And how in every version of NT, if CSRSS (Win32 subsystem runtime) crashed/returned to SMSS.exe (Subsystem/Session Manager - responsible for starting the Win32, OS2, and Posix subsystems in NT), SMSS would bug check the box immediately.
Boy, that's an interesting detail. So it didn't matter where they put Win32 stuff - if there was a crash, the box was going down automatically, by design. The Win32 -> kernel move was a big perf win (apparently), and NT has only gotten more and more stable over time, not less so.
(for those interested, this explanation appears in the grey box starting on pg 53 of the book)
When i think about this story, any of those litmus test questions would have caught it. Did i really understand the problem well enough to be talking about it? No. Had Win32 crashing been my personal experience? No - I'd refused to even use NT up until that point in my life. Why was I even bringing this up? To sound intelligent. To try and be provocative. Because Microsoft was "the enemy".
Now, when I talk about this story, I make sure to attribute the "knowledge" to the source I got it from (The "Book"). Infact, I even pulled it out while writing this to make sure I got the details acceptably close. I still don't have any personal experience with CSRSS crashing. But I "buy" the argument presented in the book until I learn otherwise (by hearing a better argument).