scobleizer wrote:Beer: I'll try to do better.
A lot of times in interviews I'm just trying to mentally keep up with the interviewee.
She's a developer on the .NET CLR team. I'm nowhere near as smart as she is.
While I understand its much more effort to arrange, I find the videos where you get the Guest and someone the Guest knows (manager or whatever) together, there is less chance for the silent parts where your thinking whats the next smart question, as the other person who knows the subject area you are talking about with the guest can then chime in. There's many videos already where we have seen this work best I think, like
Kevin Schofield - Tour of Microsoft Research
Herb Sutter, the Future of Visual C++
Kit George - Tour of .NET CLR Base Class Library Team (Part II)
This tour with expert + a good conversation with guests format is very good as there can both be expert conversation and novice questions and less chance for akward silence. And the guest may even feel more comfortable when theres someone (s)he knows better to help out if the question is going a bit out of his/her area of knowledge etc..
And my hidden agenda: Less of The Scobleizer laughing and more of the interesting conversation between people who know what they're talking about - if they go too much rocket science, you can always let them know.
But actually I feel stupid whining about this, most of the videos lately have been like I've hoped here, for this video I guess you just could not arrange the better setup?
Feb 08, 2005 at 12:49 PMWhat kind of work MS is doing on making it more easy to write more robust code that will make use of the multi-core computers of tomorrow? There's a lot of talk about the pitfalls of concurrency today and the message seems to be that even if you're a C# programmer you have to know a ton of very low level stuff and you still are likely to get it wrong or play (lock) too safe and get no benefits.
I understand that the C++ compiler has some sort of flag that makes the compiler take use of MP on marked code. It would be great if you could give the compiler a better sense of what your up to and have it figure the best places to lock and what not ..
I have some questions to Microsoft:
Why it is so difficult to create even a simple search that is _fast_ (and silent) for finding files and folders by their name. I do not know about you but most of the time I need to find a file or folder by name in the current system - It is very rare that I need to find certain string inside a document. I know I am an exception here, but when doing so much simpler search (fuzzy match string againt filename/foldername string) one would expect the local search to be atleast a somewhat faster than Google query over the network when a computer has only less than million files/folders - certainly something very much achievable by average home computer if its well done.
I know MS provides the Indexing service, but I've tried to figure it out many times, and it has never made any sense to me (I doubt it's necessary to point out where it has failed). And I am not even interested to build my own UI to the indexing service for the reason that it makes a lot of noise during the indexing (Unless you happen to have solid state storage).
I would like to present how simple it is, in theory, to make a silent full indexing:
1. Suppose user introduces a medium not previously indexed (cd/dvd/serial-ata hdd?)
2. As the disc is recognized, it is matched against previous media signatures
3. User loads a file from the disc
3a. The loaded file is entered into the index, which is partially stored in memory to avoid noisy random HDD access and lazily updated with transactions to system HDD
3b. As the user mode process does something with the file(s), in the file(s) content is being analyzed in the background (like your virus scanner does) and proper records created into the memory for the lazy noiseless update.
4. User is done with the file and does no further access in the so far mostly unindexed media.
4b. Incase the medium is CD/DVD or other very noisy one, the appropriate driver is notified to spin down the speed enough for the medium not to make noise during read (in case of CD drive this should be either user or manufacturer definable rotating speed)
4c. The last (before idle/after lowering speed) position of the hdd/dvd is picked up.
5. Now the full indexing can begin from the position of the disc, making no noise that would indicate the media is being indexed.
Certainly there are other questions, like what if user removes the media before its indexed etc - but the key point here is the silent, pleasant way of access - some of which NCQ may help slightly, but in the case of jumping around between file data and MFT - I think not.
One thing that really amazes here, is that Anti-Virus programs pretty much do everything that would be needed to create fast searchable index.. Why they do not take advantage of this, must be because searching the user file contents for certain data is not their core business (Or is it, I am not the expert here).
I can understand that some people may have not yet learned to appreciate more silent environment, but looking at what people say about the new set-top-boxes and other livingroom "PC" equipment with HDD's, you can expect that if they get a Longhorn Embedded on their box and it has noisy indexing, there may not be many positive reviews in that regard. EndRant();
PS. To anyone who cares, certain Seagate SATA HDD's may have annoying feature that writes some data to the HDD that only Seagate can read at their service during idle moments. So incase your HDD will fail, Seagate may have better clue why their certain products <beep>. Hitachi has same feature, but where the Seagate noise is like short ranged constant seeking, Hitachi noise is like the drive is going to fail any moment. The takeaway - Do Not Buy if the computer with these HDD's is anywhere near you (<10 meters).
Probably not, sorry. We tried... However, there WILL be some other kernel heavyweights coming to theatre near you
Uh, now I do not get it. Wasn't channel9 pretty much bumping to people around the places without asking ahead? Just go to where the big shots are and run into them. They won't get the chance to say no =)
Charles: This goes bit offtopic, but what kind of servers host these sites (blogs.msdn.com channel9).
Should world go 90% or more underwater tomorrow - would the survivors a) have a backup located in place highly unlikely to go underwater b) if they attempted to rescue MS websites and Longhorn, would that be possible without flooding the place where the information is stored (the datacenter would need to be accessible underwater without difficulty). c) Perhaps there should be a floating satellite link and satellites which can store all the data currently on the ground.
Am I serious? Well good question, hard to answer. Recently there has been so many floods, even HERE where never to my knowledge has been such floods, that thinking these kind of possibilities cannot be entirely avoided. And I love watching those serious looking documentaries that tell about disasters and show nice graphs to support them - I have no time to go do background research of the subjects unfortunately.
My wild imagination brings me the thought of a city designed to survive the waves and then float. Like houses seen on some recent news footage. Internet would still need to work underwater - but that it does already!
Oh and while this is nothing compared to what you see on TV, here is some pics I managed to find..
edit: Oh and needless to say, I loved these videos, but you knew that already Right?!
I think the main reason why Apple seems to kick in the OS area is that they are not so big in backwards compatibility. For MS it is much harder to make big "innovations" (talking about the GUI) happen as it is not acceptable to just start from scratch every now and then. Though atm I have to wonder what kind of spaghetti the OS X code is to have enabled some of the features they demoed. Maybe they'll spend more time in the future refactoring instead of innovating, unless they scrap everything again - which I doubt a bit this time.
BTW Am I the only one kicking myself for not getting some Apple stock before it rose back to the '00 level Lets hope there is a time again where MS stock is in demand
(technically OS X was not started from scratch of course, but I'll limit this to the GUI. MS however is surely carrying some _very_ legacy code even when talking about running the GUI of LH)
Hey Josh, if you happen to be watching the thread - do you know did Mark Cligget move onto other tasks or what, I remember him being very active towards the VS community on his blogs, but then went suddenly silent about a month before beta 1.
OT: Using the Save link to download the video, seeking doesn't work in the saved wmv and it stops playing at some 75%.. These problems do not appear in the streaming version.
OT2: I have had ch9 since beginning using cookie for password - I do not remember the password and cannot access the profile from other computer. Can you please send a new password to the email in my profile? I tried the forgotten password function but it does not seem to work in this case. Thanks.
Is the "sorting" going to be easy to define? For instance I have only few albums, mostly singles and ep's etc and I sort them by the record label as its often more defining that genre sorting. Also suppose you have a SQL server with the data, can the devices get the "views" from there too?