共用方式為

You’re asking for the wrong thing.

  • 發行項

Just stumbled upon this article by John Dvorak titled “My Windows 7 Wish List”, and I see the same pattern as many of the other fruitless wish list articles. Maybe IHBT…

As a user, the user experience (UX) is the only thing that ever matters to you – the implementation is never the issue.

Take for example item (1) on the list:

“Build a new file system based on database technologies” . Why do you care about how the file system works internally? Do you actually want a file system with a DB in the backend, or do you just want to be able to run fast, useful searches? My guess is that it’s the latter.

Another one:

“Get rid of the miserable registry” . Is it the registry you care about, or the fact that it’s difficult to upgrade/migrate?

Another:

“Get off the cloud” . You’re worried that “suddenly the machine wants to contact the Internet for some reason or other”. This has nothing to do with the “cloud strategy”; you just want a machine that doesn’t use the network. Or something. I don’t really know based on that rant :)

 

Alternatively just read this:

https://shippingseven.blogspot.com/2008/07/20-features-windows-7-should-include.html

 

Basically (and I’m talking about customers in general – not just MS/Windows) people often request implementation changes, but what they really want are UX changes. Do you want a powerful engine, or just a fast car? Do you want high carbon steel, or just a knife that’s very sharp and stays that way? Do you really care how the sausage is made? ;)

 

As a customer, it’s important to understand what you actually want since it makes it much easier to ask for it (and hence get it).

 

Avi

Comments

  • Anonymous
    October 15, 2008
    Since you brought it up: Get rid of the miserable excuse for a file system that is NTFS. In terms of performance, it's one of the slowest systems I had the displeasure to work with.What I want is Microsoft licensing ZFS and offering that. I don't want you to reinvent a file system - FAT16, FAT32, NTFS, and Samba prove that it's not one of MS's fortes.So yes, many of us want implementation changes, based on the knowledge that pretty much anything would be better than what we have now.I'm sure that's not what you wanted to hear, but there it is... ;)As for the searches - if you offer them in the FS, can you please not invent a new technology to query them? How many query languages/frameworks did MS go through now? Can we stick with one?Even better, can we have an open, well-documented standard?
  • Anonymous
    October 15, 2008
    Wait, you want ZFS, or you want a file system with features X/Y/Z (including more speed)?New technology for searches - why do you care what tech is used under the covers, as long as the search facilities do what you want them to?I'm not arguing about the merits of any given feature; I'm just saying that you're requesting something irrelevant (ie; the implementation, rather than the end result).
  • Anonymous
    October 15, 2008
    The implementation is actually extremely relevant as soon as I want to go cross-platform. So yes, I want ZFS - it's available on all platforms except Windows and has the features I need. I don't want MS to attempt to ape it. If there's a better FS coming soon, maybe - but even then there's the cross platform issue.And yhat's why I want to use an existing query language for search. I don't want to spend time working around quirks and learning new APIs just because Microsoft thinks they should reinvent everything.There's plenty of cool tech coming out of MS, but if there's an open alternative that is at least on par, it might be worth using.The request that open standards and open source be used is a feature request. It's a pity Microsoft doesn't seem to get that. (Or is getting it very slowly)
  • Anonymous
    October 15, 2008
    >The request that open standards and open source be used is a feature request<<Yes, that's a legit user request; it doesn't discuss implementation. >I want to use an existing query language for search<<Also a good feature request - it would apply regardless of what the backend implementation (RDBMS, file-based index, etc) is.I'm not disagreeing with your points (ie; you want MS to do some things differently); and I'm not totally certain you're disagreeing with mine (implementation isn't the user's interest), either :)
  • Anonymous
    October 15, 2008
    Bravo! Good points! :-)So, when you plan to improve searching perceived performance? If the REAL question is so obvious and is known for such LARGE amount of time, then WHY you have not improved this?;-)
  • Anonymous
    October 15, 2008
    “Get rid of the miserable registry”. Is it the registry you care about, or the fact that it’s difficult to upgrade/migrate?Actually, I can tell that I for one care about that. I do not want "yet another thing" when there isn't need for one. Then again, if you can take registry and wrap it transparently so that it looks as a usual FS folder with files under some "magical" path in the FS proper (say, C:WINDOWSRegistry), so that I can use all the usual file tools to deal with it (including upgrading/migrating), I'll be fine with that. I do not want another interface/API to deal with, I don't really care about implementation.Also, when people say they want a database-backed FS, they usually mean that they want transactions and relational-style queries (preferrably SQL - why learn a new language?) for their FS, with performance comparable to that of a proper RDBMS. If you can pull that off without hoisting FS on top of MSSQL, sure, go for it (but wouldn't you end up just writing your own database in the end?).
  • Anonymous
    October 16, 2008
    Wojciech, I'm afraid I don't work on the Windows team, or even close to it. I have zero power to make changes there. And I'm pretty sure I haven't suggested that there's a known/obvious way to make anything go fast[er].int19h - good call about a FS interface to the registry; that would be an interesting migration path! As for relational style queries on the FS; LINQ shows us that this is possible regardless of the backing store. Performance of an RDBMS: SQL Server uses indexes just like Vista uses an index for search - the difference is that SQL Server's whole reason for existing is query speed, hence it can make optimizations for that (such as keeping as much of the index in RAM as possible) that Vista cannot.Either way, seems we've veered way off the original post topic now :)