Forum Posts Feed
Web Research -->> WOW!
< Next Topic | Back to topic list | Previous Topic >
Posted by Graham Rhind
Jul 5, 2007 at 04:05 PM
quant wrote:
>>This doesn’t make sense to me. I don’t know how UR works internally,
>>but adding a
>record containing (for example) a web page shouldn’t require any other
>>records,
>whether they contain metadata or not, to be updated. In my database program
>>of
>choice (Visual Foxpro) it is as quick to add a record to a database containing 2
>
>>million records as to one containing 10 records.
>
>I’m not a database specialist,
>does the speed of insert operation depend on whether the database is kept indexed or
>not?
“Insert” commands shouldn ‘t physically insert a record (i.e., pushing all the records below it downwards) - that would kill off any use a database may have had as it is excrutiatingly slow. Insert just adds the record to the end of the file. Inserting into an indexed (ordered) file will affect speed because the positioning of the record in the interface needs refreshing.
Posted by quant
Jul 5, 2007 at 04:43 PM
Graham Rhind wrote:
>... Inserting into an indexed (ordered) file will affect speed
>because the positioning of the record in the interface needs refreshing.
UR keeps all the keywords indexed, so you answered yourself :)
Posted by Derek Cornish
Jul 5, 2007 at 05:48 PM
JJ,
I’d guess a lot of the UR users would claim that its information management abilities extend far beyond those of WR, and approach those of Zoot (this from a Zoot-biased viewpoint, needless to say). This would explain their strong resistance to the idea that WR could be a replacement for UR. But of course it all depends on what one is using UR for. Being a WR user myself, I can see how its developing feature-set might make it an attractive alternative to UR or Zoot as a single-product HQ for some users (though clearly not for others!). When using two decent programs with overlapping features one is constantly alert for the tipping-point at which one’s usage might be switched to one or other of them, rather than both.
Not being a heavy UR user I would not want to get into any WR-UR tipping-point discussion, particularly as it depends so much on an individual’s usage. But to take a somewhat analogous example, WR certainly does not have the feature-set to replace my use of Zoot at the moment, although I have often wished it did. It is, however, a very useful companion program to Zoot, in the latter’s current pre-32 bit state. This is why I stressed the ability of WR to provide hyperlinks to its content that can be pasted into external programs - something the Admiral kindly ensured that Zoot could handle. Currently in WR the hyperlink is only to the saved file, not to its contents (paragraphs, sentences, words, etc). But this is still extremely useful if one wants to use WR as a well-organised repository for webpages, images, and so on, when used in conjunction with Zoot or other similar programs. (Incidentally, I am at something of a loss to understand why this feature is so rarely commented on when discussing WR, especially as it has been praised here in relation to Whizfolders: see http://www.outlinersoftware.com/messages/viewm/1833.)
But the point I was really trying to make - not very well - in previous posts, was that it would be a pity if the good points of WR as content-capturing software, in comparison with other software in that class, were to get overlooked in the scramble to identify its shortcomings with respect to UR or other more powerful information organisation and management software. Like you, I invite people who have not d/l WR - especially those who have commented on it :-) - to take a look.
Derek
Posted by Graham Rhind
Jul 5, 2007 at 05:52 PM
quant wrote:
>UR
>keeps all the keywords indexed, so you answered yourself :)
I don’t recall asking a question. But does this mean that UR scans incoming web pages for occurrences of keywords? That would certainly explain why it is slower than Web Research in jj’s test.
There is a difference in the speed that programs capture web pages - I find Surfulator very slow on my machine, for example, where no keywords are involved. After a quick look at Web Research I have to say it is lightning fast. I am still installing the network add on, which will give it the same indexing burden as UR. jj has already tested this and suggests that it is faster, which I can well believe. It also seems to be faster in file importing than UR, but I haven’t tried this yet with the network add in.
Posted by JJ
Jul 5, 2007 at 05:59 PM
Derek Cornish wrote:
>JJ,
>
>I’d guess a lot of the UR users would claim that its information management
>abilities extend far beyond those of WR, and approach those of Zoot (this from a
>Zoot-biased viewpoint, needless to say). This would explain their strong
>resistance to the idea that WR could be a replacement for UR. But of course it all
>depends on what one is using UR for. Being a WR user myself, I can see how its developing
>feature-set might make it an attractive alternative to UR or Zoot as a single-product
>HQ for some users (though clearly not for others!). When using two decent programs
>with overlapping features one is constantly alert for the tipping-point at which
>one’s usage might be switched to one or other of them, rather than both.
>
>Not being a
>heavy UR user I would not want to get into any WR-UR tipping-point discussion,
>particularly as it depends so much on an individual’s usage. But to take a somewhat
>analogous example, WR certainly does not have the feature-set to replace my use of
>Zoot at the moment, although I have often wished it did. It is, however, a very useful
>companion program to Zoot, in the latter’s current pre-32 bit state. This is why I
>stressed the ability of WR to provide hyperlinks to its content that can be pasted into
>external programs - something the Admiral kindly ensured that Zoot could handle.
>Currently in WR the hyperlink is only to the saved file, not to its contents
>(paragraphs, sentences, words, etc). But this is still extremely useful if one wants
>to use WR as a well-organised repository for webpages, images, and so on, when used in
>conjunction with Zoot or other similar programs. (Incidentally, I am at something of
>a loss to understand why this feature is so rarely commented on when discussing WR,
>especially as it has been praised here in relation to Whizfolders: see
>http://www.outlinersoftware.com/messages/viewm/1833.)
>
>But the point I was
>really trying to make - not very well - in previous posts, was that it would be a pity if
>the good points of WR as content-capturing software, in comparison with other
>software in that class, were to get overlooked in the scramble to identify its
>shortcomings with respect to UR or other more powerful information organisation and
>management software. Like you, I invite people who have not d/l WR - especially those
>who have commented on it :-) - to take a look.
>
>Derek
Derek,
I couldn’t have said it better :-)
Have you tried using WR with the network add-in???
PS… I too am looking forward to Zoot32!
-jj