Found a couple interesting links over at NooFace, a site dedicated to user interfaces. First, Microsoft Research is plugging away at one of the growing dilemmas in computing: so much data, so little time. XTend is a next generation product which uses a powerful Relational File System to deliver the world’s first “Save With” oriented interface. And speaking of Knowledge Management (where XTend can be useful), you can read this recent interview over at TheIdeaBasket with the CEO of YellowPen, Inc.
Microsoft Research Focusing on Search Interfaces
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
If they are so focused on improving search interfaces, maybe they could kill that stupid little “XP dog”? I’d prefer if it was shot Fallout-style (those who played it know what I am speaking about).
Looks like the already massively bloated WinAPI will become somewhat more of a behemoth. Longhorn looks to be heading toward needing 2 cds.
Win32 is a system interface, it’s more likely that specialized libraries of vendor research and development will be added to their user mode object-oriented platform (.Net).
Not sure what I think about the articles. It’s not like I have all this personal information that I have to organize. Sounds more like a business need. I only like the idea of using it to block out information like spam. Can’t wait to use all of this on Linux.
This idea is interesting
But, can’t this essiently be done with the traditional folder heirachy, make a folder for each client/project/whatever.
The other thing I don’t get (maybe I didn’t dig hard enough though) is that sure its saved with the other documents you select, but don’t you then need to remember where the other documents are?
Also is going to the effort of tagging each document with each other related one really any simpler than just remembering a filename?
To me this just seems to be sidestepping a problem and introducing others.
I think there does need to be a new method for storing documents so that it makes it easier for less technically minded people to use a computer, but I’m not convinced this is it.
7 CD’s or one big fat fully loaded DVD.
What is so hard about:
1) naming files logically
2) placing files into folders
Maybe it is about time started learning to use computers rather that side stepping the issue and using the old blame game of “computers are too hard”.
Anyone see the ad for “Tony Robin, “lifes not boring, you are”, apply that to computers, “computers are complicated, You’re just as thick as two short planks!”.
Lets see how long it takes until they (MS) buy the Xtend technology or the whole company and then say, he look we have been innovative again!
“Not sure what I think about the articles. It’s not like I have all this personal information that I have to organize. Sounds more like a business need. I only like the idea of using it to block out information like spam. Can’t wait to use all of this on Linux. ”
How many people out there have big drives, and plan to soon add more? Even if you’re fanatical in your organization. things are going to be, at least slow to find, at best it may be lost. It should never be amazing how much information even an individual can accumilate, and how quickly too.
“What is so hard about:
1) naming files logically
2) placing files into folders
Maybe it is about time started learning to use computers rather that side stepping the issue and using the old blame game of “computers are too hard”. ”
Well on general principles I would like to have a more educated public. However the problem with the above is that they “break down” when it comes to large amounts of information. That’s why databases were invented. We’re in an “information glut” coming from many sources, with no end in sight. Remember “why” computers were invented? From the first day there was nothing between bare metal and user. To the ever increasing levels of abstraction, putting more and more distance. Now another layer is needed to keep our “tools” useful. The research is needed because there’s no moore’s law for humans.
Thanks for taking the time to mention XTend as noteworthy…
Some answers to the comments so far:
Ores
Remembering what a document is saved with is a lot easier than remembering where it is. In XTend you only need to remember one thing it is saved with. And you don’t need to know where that thing is. At some point you have to know where to look for sure. But this is where a keyword search comes in.
The ‘effort’ of tagging is not really much of an effort. Because it is a simpler replacement for choosing a location to save a document. It is not additional work it is less work… the key to good UI design.
Matthew Gardiner… If you don’t understand the problem you won’t understand the solution!
The key to Save With is that you can save a document with many things and all you need to remember is one thing to recall it. This is not the same as logically naming file and putting things in folder. Cause I can guarantee that you can’t come up with a folder schema that can logically and simply capture the complexity of most businesses without introducing barriers to information and a need to remember too much about each file before you can find it again.
Take this problem.
You are company with clients and products (standard problem)
So you create a Client folder hierarchy and a Product folder hierarchy. Everything is fine until something is both about a client and about a product. Where do you put it then? One solution is to put a copy in each folder… argh version control anyone…
The other commonly used solution is to put one of the hierarchies inside the other… i.e. Client1Product1 etc.
Question: Can you see everything related to a Client1 at a glance? Answer: No. Reason: The Product subfolders act as a barrier to the information.
Question: Can you see everything related to a Product1 at a glance? Answer: No. Reason: The product information is spread throughout all the various client folder.
Of course in real life there are usually a lot more than two dimensions to data…..
Also why should people learn work like a computer? Unlike computers we don’t have perfect memory and the ability to handle great quantities of information.
Regards Alex James
http://www.maxtiviti.com
Sir, “Save With” is a fine idea, I only would suggest providing the User Guide in a more portable format like PDF.
[ http://www.maxtiviti.com/assets/marketing/documents/The%20XTend… ]
I’m using Microsoft WindowsXP, but I don’t have Microsoft Office installed, so the DOC file is opened in lame MS-Wordpad 5.1 . And that happens after a cumbersome Microsoft Word 97 Conversion:
“Unable to load graphics conversion filter. Continue with document conversion?”
I have to click -YES- twentythree times before the glorious Guide finally opens.
I’m enjoying XTend, thank you.
It opens in OpenOfficeWriter on Linux.
ahhh the joy of linuz. forgot to remind that the doc obviously gets opened with no screen shots (that is if u don’t have super Office). (;>J)
Sorry “m”
We are currently working on some flash movies which should make reading a guide obsolete. Refer back to the http://www.extend.co.nz/Demos.htm in a couple of days and I should have the first of the new demos up.
I would love to hear any feedback you might have on XTend.
Regards,
Alex
I am Alex James and our email format is [email protected].
Alex James
Founder
Maxtiviti Solutions Ltd.
http://www.maxtiviti.com
What I wouldn’t mind seeing though is a distributed public knowledge base. A virtual generic framework that can be reused by individual non-corporate service providers. Provide people with the tools to become involved in markets that are closed due to the nature of information technology: the advanced educational requirments, the dependance on large scale teams, access priveledge to closed implementation, knowledge and control through democratic evolution of architecture.
I’ve seen the X-tend filter thing in Konqueror, at least…
The concept is nothing new…
author – subject – context – date . suffix
for example, I am completing a paper regarding a programming in 4GL language
MatthewGardiner-4GLProgramming-howto-270403.doc
That would be placed under /home/matgarnz/documents/assignment/programming
Regarding the senario with the customer, you’d place all the information regarding the customer under one folder then expand out of the that folder
~/customers/[customername] <- top level for the customer
~/customers/[customername]/[context] <- Context for the customer, for example, if it is about buying books, then the directory would be called BookPurchase
Jonas, what were you smoking then?, better give a link to what you saw in Konqui or move on.
Mathew, like it or not, some people [certainly not all] are out to stop -visually- slashing things, searching for alternative ways of retrieving and tagging information.
Call it crazy [not me]. You sound like fifty four years ago telling everyone that IBM punch cards are just not going to work, “naaah, look this is how I keep my accounts, I put costumers here and context there…”
Abstraction works, heck that’s the essence of computing, you can put everything in one single place and retrieve what you are looking for by its tagged metadata. That is a more neat context than endless slashing.
[ http://www.columbia.edu/acis/history/punch.html ]
I can’t help comparing this -Xtend- ‘search tool’ with Gelernter’s -Scopeware Vision-
[ http://www.scopeware.com/products/prod_v_individuals.html ; original concept: http://www.acm.org/sigchi/chi96/proceedings/videos/Fertig/etf.htm ]
Both systems try to elude the current hierarchical management of documents, -Scopeware Vision- retrieves files in a time-centric way (a time-ordered stream) and -Xtend- seems to deploy a category-centric model. I haven’t played much with -Scopeware- (the former beta had an indexing problem with my system), I don’t remember if/how docs got categorized, gonna download it again.
I believe Microsoft is implementing the essence of this project in their new filesystem which should be released with longhorn. Interresting article at news.com:
http://news.com.com/2009-1017-857509.html
In the grand scheme of things, sure, it sounds nice, however, I do think there are more important things to worry about that trying to create new gadets to help disorganised people.
How about self diagnosing and healing operating systems? when the OS crashes, on the reboot, it analyses the memory dump to work out the cause, say, if in the case of bad memory, the operating system crashed, the operating system would then learn not to use that section of memory again, and stop any applications from accessing it as well. That is the kind of stuff that would be useful to productivity increases.
“Take this problem.
You are company with clients and products (standard problem)
So you create a Client folder hierarchy and a Product folder hierarchy. Everything is fine until something is both about a client and about a product. Where do you put it then? One solution is to put a copy in each folder… argh version control anyone…
The other commonly used solution is to put one of the hierarchies inside the other… i.e. Client1Product1 etc.”
You put it in product and make a hardlink to it in the client dir.
“Question: Can you see everything related to a Client1 at a glance?”
Yeah, in this example you’d have a link to the product in his directory.
“Question: Can you see everything related to a Product1 at a glance?”
Well the product should be in or at least have its own directory so then you could have other related stuff within it, links to manufacturer or whatever.
“Also why should people learn work like a computer?
Of course there’s limits to what you can do easily but most of what you might want is already possible and could be made quite straight forward given a good interface.
10 years ago I read a glowing magazine article about the talented geniuses working at Microsoft Research. The article predicted that, well about now, we’d be seeing the fruits of their endeavours. So, have we seen any novel ideas from MS Research reflected in the products available today, or has MS used this part of their company to simply buy off researchers so they cannot work for other innovative companies? As far as I know, MS Research is only a shadow of the types of groups they tried to emulate (i.e. Xerox Parc).
“Lets see how long it takes until they (MS) buy the Xtend technology or the whole company and then say, he look we have been innovative again!”
MS wouldn’t do that. They would copy the idea, bundle it into the OS, put Yellowpen out of business, THEN call it innovation
> Jonas, what were you smoking then?, better give a link to what you saw in Konqui or move on.
Can’t give you a link but there’s a filter in the extra bar that shows only files of some specified types in one directory (I guess it could filter by other criteria either).
But I missed the point!
The concept is may be well beyond that (except for one directory you have an entire system).
Noneless I tried the tool and think the whole system sucks.
(Same goes for Scopeware). The reason is simple: Use KISS (Keep It Simple Stupid).
So I’ll go back to my Linux/*nix boxes…
interesting that you mention this R&D facility. The only thing I can recall being developed by them, and incorporating into a product is the Microsoft implementation of TCP/IP version 6, which became “production quality” just recently and incorporated into Windows 2003 Server.
As for the rest. I would like to see something really innovative come out. Something that would really knock my socks off, just as Xerox did with the GUI, and digital did with the Alpha chip. Both companies I must note did not have the money Microsoft has now, I so actually expect alot more from Microsoft.
> Jonas, what were you smoking then?, better give a link to what you saw in Konqui or move on.
Can’t give you a link but there’s a filter in the extra bar that shows only files of some specified types in one directory (I guess it could filter by other criteria either).
But I missed the point!
The concept is may be well beyond that (except for each directory you have an entire system).
Noneless I tried the tool and think the whole system sucks.
(Same goes for Scopeware). The reason is simple: Use KISS (Keep It Simple Stupid).
So I’ll go back to my Linux/*nix boxes…
Everybody criticizes MS, but only I get moderated down? And at the *very* poor excuse that I would be attacking this site?
At least use one criterion!
This kind of “need” can be solved by existing systems that work well, but are rather unused. And I don’t mean anything in NTFS or FAT32 or HFS+. You know what I’m talking about…
I have to agree with Matthew Gardiner when he asked “why is it so hard for people to give meaningful names to files and to place them in folders??” It’s true. It is a basic thing that should be step one in all basic computer skills classes (instead of jumping right into “Using MS Office”). The sad fact is that people are generally disorganized and lazy (this isn’t some kind of anti-user bashing; I mean people in general in all areas of life are disorganized and think that it takes too much time and effort to be organized, which it does not). I see this all over the place, not just on people’s computers. It’s an age-old human attitude problem that leaves the way open for all kinds of products to be marketed towards people who think that a book of tips or a new computer software package will MAKE them be organized people. None of these things will do anything more than give you good ideas. When people don’t have the organizational HABITS, good ideas and great organization systems are useless.
There was an interview with someone from M$ this week saying that there will be a CLI Windows Server …
This is the future of Windows’s interface 🙂 hahahahahahahaha
This is the URL for the interview:
http://news.zdnet.co.uk/story/0,,t269-s2133899,00.html
CLI is still the best interface for servers. Unix and Linux was rigth 🙂
Jace…
It is true that some people are truely disorganised, and no amount of cool tools will help them.
XTend http://www.maxtiviti.com/XTend.htm is not claiming to magically make things better for these people.
The problem is no matter how organised your files are storage becomes more complicated overtime. I.e. moves from one dimensional (Client folders) to 3 or more dimensional (Client, Deal, Product etc. etc) and the current systems we use (i.e. the Hierarchical Systems) stop paying divideds.
I used to find that my thinking was “Where can I put this so I won’t lose it?”. When placing a file deep into a nested hierarchies you are actually hiding the information, not making it easy to find again.
Unless you remember the Client, The Deal and the Product you won’t find the document in the folder
Client1Deal1Product1
Not only that you have to remember all these things in the correct order.
Why? Because until you go into a categorising folder it obscures what you are looking for. So the more categorising folders you choose the harder it is to find information. Which is the exact opposite of what you are trying to achieve by putting the information into many categories.
XTend is significantly different, as you only need to remember one dimension to the data (say the Product) and in theory you should see your file, if there are lots of files then the other categories assigned to the file don’t act as barriers, the act as prompts. I.e. the file is already visible, it just becomes MORE visible when you filter on the next thing too.
I.e. the file is visible here:
Product1
the file is MORE visible here:
Product1Client1
the file is MOST visible here:
Product1Client1Deal1
or
Client1Product1Deal1
or
etc….
This is not the case with a Hierarchical system… where the file is INVISIBLE everywhere except the exact ordered combination of categories… i.e. you must be in Client1Deal1Product1 or you wont see it!
See this is not about being disorganised, it is about having a system that allows people to capture all the meta data about their information without creating completely unwieldy structures.
For more information http://www.maxtiviti.com/XTend.htm
Regards
Alex James
Founder
Maxtiviti Solutions Ltd.
I’m not impressed.
If you don’t know where the file you need is anymore, there’s always find(1) (or the less powerful windows-equivalent).
While it sure is a nice idea, it’s just a gadget imho. But then again, that’s what MS is all about: making an OS full of gadgets, because they make nice selling points
Can’t blame them for wanting to run a succesful business, can we?
The problem with Find/Search is it works on Name, Content and/or location. If you find a useful document somewhere but later want to bring it up with all the files for a project you are currently working on (lets say it’s an Example) unless the current content or current path mentions your Project it is impossible.
Adding a relationships means you can effectively alter how something is found, without moving or modifying it.
Sorry your not impressed ;(
Have you tried it?
You might be surprised.
Forgot to mention the other issue File Consolidation.
Suppose you do remember where 4 ‘Product’ files are
They are:
/Client1/Product4/File1
/Client5/Product4/File2
/Client3/Deal2/Product4/File3
/Product4/Manual/File4
How do you quickly work with all 4? Open 4 windows? See all the other irrelevant files in each of those folders?
XTend can consolidate information from all over the place into one VIRTUAL location. Eliminating the need for manual consolidation.
Regards
Alex
You know, this isn’t THAT new… with BeOS, I can store things and give them some tags (“attributes”; apps usually do that automatically) and run a VERY efficient search.
I remember searching for a very specific e-mail.. I was surprised I had misplaced it, but found it in no time.
I’m also impressed at BeOS’s search speed… ALL results are found (no matter how many folders deep they are) before 15 seconds…
You know, I like the Office Assistants.. that’s the only thing I miss in OpenOffice…
If you don’t like the dog, well, you can turn it off. It isn’t that hard, you know?
15 seconds? I don’t think I’ve ever had a query that has taken that long to complete. It usually shows up instantly.
Attributes is great for describing files, even though it can be a bit tiresome to put all the meta-data there unless the process is automated somehow. It’s surprising that it has taken this long for MS to implement something like this considering that their OS is widely used at companies(who really needs this kind of stuff).
I wonder if they will manage to create a good interface or if they will screw it up with MS-userfriendlyness as usual.
So far BeOS has the best implementaton of an attributed filesystem (from the users perspective) that I’ve seen. But It’s far from perfect. So here’s your chance to shine MS
This sounds like a crufty solution for a situation where a wee bit of common sense and group effort would yield far better results.
At the company I work for, our approach to our information infrastructure is something like this: We organize our files in terms of clients, because that’s who we’re ultimately working for.
On our file server, there is a “Clients” folder. Within that folder, there are folders labelled A through Z. Each client has its own folder within the lettered folders based upon the first letter in their name. Every file related to a particular assignment for a client goes in the folder.
So, if you need to work on something for Client “XYZ, Inc.”, you’ll find everything under /clients/X/”XYZ, Inc.” When it’s all said and done, there is *zero* ambiguity as to where a file is or what is its purpose when you approach things from the perspective of who your client is.
Look at it this way…before we had computers, this was absolutely the fastest way to store all our data…in real, metal file cabinets and manila folders.
Thanks for the response. I appreciate the examples. I do also appreciate the system’s merits, but… this system is not needed on an OS like the one I run, for reasons mentioned by Lanjoe9, above. The attributes and queries in BeOS really do provide this kind of service. In the end, it all boils down to the same issue again: people need to use meaningful names and be organized in order for any system to benefit them. I’d be happy if Apple and/or Microsoft adopted the BeOS methods for attributes, queries and (last but not least) file type management. It would make those operating systems more functional to me when I have to use them.
FYI
A new flash movie introduction to searching with XTend http://www.maxtiviti.com/assets/marketing/search.htm has been put on our website and is particularly useful I hope if gives you all an idea of how searching in XTend works.. Might helps some people visualise what I have been writing about.
Regards
Alex James (email = [email protected])
http://www.maxtiviti.com