Greetings.

I am looking for a PIM that works on a MAC, but features multi-database searching (and perhaps highlighting of the hits). I am currently using UltraRecall, a great program, but it lacks multi-database searching and highlighting of hits, though these features are on the roadmap.

As soon as I get my MAC (I am fed up to here with VISTA) I will take a look at DEVONthink, but I can't find anything on the Forum or on the company's webpage whether or not the above features (i.e., seaching across databases, when the data becomes to large and has to be broken into multi-databases) are available. I would deeply appreciate your advice.

Not entirely sure what you mean by this. Opal, frex, is Spotlight searchable, so you can search multiple documents that way. Its filtering searches within a single file, showing you all relevant topics.

Many if not most Mac PIMs support Spotlight.

OK. Let me clarify, using UltraRecall as the PIM, albeit a non-MAC PIM.

Sometimes a certain UR database becomes "too large." So inevitably I would have to create another UR database to hold new data, so now I have two very large UR databases. I want to find certain data, so I need a PIM with a search function that operates across databases (in this case only two). Currently, UR can't do this, though, as I said, it is in kinook's roadmap. Is this clear enough?

Yes, DevonThink can search across multiple databases.

basilides wrote:

OK. Let me clarify, using UltraRecall as the PIM, albeit a non-MAC PIM.

Sometimes a
certain UR database becomes "too large." So inevitably I would have to create another
UR database to hold new data, so now I have two very large UR databases. I want to find
certain data, so I need a PIM with a search function that operates across databases (in
this case only two). Currently, UR can't do this, though, as I said, it is in kinook's
roadmap. Is this clear enough?

To be precise:

Devonthink Professional and Devonthink Pro Office - both in version 2 - allow you to open and search multiple databases.
Devonthink Personal and Devonthink Note do not.

See http://www.devon-technologies.com/products/devonthink/comparison.html

Franz

Franz, if I may,

Thanks for the information. By the way, do you know when DevonThink 2.0 will be out of beta? I couldn't find any indication on their website. It seems to be the program I am looking for, at least until UltraRecall incorporates the two features I mentioned and a MAC version is made available.

It's unclear when it will be out of beta... my guess is another six months. But it's usable now (and has been usable for quite a while; two features are not yet implemented however). One of the major focuses in 2.0 was improvements in the multi-database interface. The latest beta has a nice "inbox" as well where you can drop things quickly before sorting into individual databases.

-- Chris

basilides wrote:

Franz, if I may,

Thanks for the information. By the way, do you know when DevonThink
2.0 will be out of beta? I couldn't find any indication on their website. It seems to be
the program I am looking for, at least until UltraRecall incorporates the two
features I mentioned and a MAC version is made available.

Basilides.

Sorry, I don't know when it will be final. I am not even sure the developers know.

The current beta seems to be stable, I haven't heard complaints about crashes. But to be honest: Right at the moment, I do not use Devonthink as my main work has to be done on Windows at the moment.

Franz

A clarification, please. According to Stephen Z. and Franz G., DevonThink Pro features searches across databases, but Chris T. says the two features I am interesting in (global searching and hit highlighting) are not yet available in 2.0 beta. So who is correct here?

basilides wrote:

A clarification, please. According to Stephen Z. and Franz G., DevonThink Pro
features searches across databases, but Chris T. says the two features I am
interesting in (global searching and hit highlighting) are not yet available in 2.0
beta. So who is correct here?

basilides, all are correct. The features you're seeking are already available in the betas of DT Pro 2.0 and Pro Office 2.0. Two other features - possibly more - aren't available yet. (I suspect one missing feature Chris was referring to is full tagging functionality.) Anyway, it sounds as if DT fits your requirements, so it may be worth giving it a trial.

H

The two main missing features I was referring to were full tagging (right now it supports tagging in the same way DT 1.0 did) and the sorter (which is not really part of core DevonThink). Global searching works.

-- Chris

basilides wrote:

A clarification, please. According to Stephen Z. and Franz G., DevonThink Pro
features searches across databases, but Chris T. says the two features I am
interesting in (global searching and hit highlighting) are not yet available in 2.0
beta. So who is correct here?

Hi basilides,

you wrote, "Sometimes a certain UR database becomes ?too large.? So inevitably I would have to create another UR database to hold new data, so now I have two very large UR databases.".

Since UR is one of the only two ( ? ) Windows outliners that rely upon an SQL database ( and the one that isn't in beta stage only ), thus promising to manage real large databases, I am electrified by your post. Could you share your experience with us, giving details of the overall size and of contents ( repartition of sizes and number of documents, formatted text mainly or web sites and / or photos e.g. ), and details of your "living with it" :

When ( size ) and why ( response times in going to a given item, and / or in searching ) got your database "too large" ? I had been hoping that UR would permit to have databases as large as 1 GB, with acceptable response times ( and perfect stability ) : Have I been dreaming of impossible things ?

Since UR isn't quite dead, so a lot of folks besides me should be as highly interested in your experience, so thanks for detailing !

UR is a great program, one that can handle huge amounts of data, but I've got several encyclopedias with thousands upon thousands of entries (nodes), so I found it to be easier to divide up the data I need stored into several UR databases. This is just an idiosyncrasy, not a UR limitation.

Thanks a lot for answering. But I'm still wondering :

Since UR has hoisting, for sheer reasons of a better overviewd, cutting a big file in half would not make much sense, and all the less so since you are not cutting a big file into several distinct to-be-published-books or something like that, but you are speaking of one big file which in its end use would be one big piece of work, not two or three.

So your not being so much happy anymore with your very big file should have some reason(s) with response times, be it in navigation ( i.e. showing of the contents in the panes after clicking on the items in the tree, or even unfolding of sub-trees ), be it in searching ( i.e. simple searches but over hundreds of MB or rather elaborated ones ( which are possible in UR )).

I think several hundred MB, spread into 100.000 or more items, should NOT be of any problem with response times on a "normal" pc in respect to these two critical factors, and if it is not, I do not understand why cutting all those things into several files would be helpful or satisfying in any way, since UR's USP IS ( or should be ) the capability of holding all your stuff in one place, thus avoiding hampering with a hundred or so different outlines ( which is my problem with ActionOutline ) -

except for declining response times or whatever, and I'm really interested in that whatever :

Since the background of my question is : Is it a good idea to switch to a SQL-based outliner, once and for all ( or let's say, under the condition your stuff will ever stay under 2 GB ), or will response time problems ( or stability problems ), occurring when you get really big stuff in your software, cause in a way the same problems you've got now, with no-SQL-based outliners, i.e. the ( more or less ) "necessity" to cut files into pieces.

Hence, my question, have you encountered slowing response times from 10.000 items to 60.000 items or more ; ditto from a file size of 10 MB to 200 or more MB ? What ARE the response times, anyway ?

You ARE aware that your kind answers would be crucial in any serious verdict upon UR ? That's why a detailed answer of yours would be so much helpful to the community...

Besides, Chuck, asking for deep links and critisizing the lack of speed of development, is right :

in general, yes, that feature should be ubiquitous by now ( item to deep-link, deep-link to deep-link and deep-link to item, and even item to item : since today, in most programs, when you do a link in your tree to another item ( in another tree or even in the same ), when you rename or delete the other item, your link is not updated but hanging... ),

and in particular if I may say, since when asked for deep links, UR developers answered, we have cloning, but deep links are too complicated to implement for us, so be happy with clones for the time being ( of course, I'm paraphrizing, but it's in their forum ).

Besides, that's one of the points I had in mind when saying, create an outliner for the masses, and they will be happy to make their employers buy it for them, but bear in mind the numerous appeals MS Works has for thoses masses : If in the outliner, "it's not there", people won't switch in numbers, but continue to hamper around with MS products.

paraphrasing, gosh, ugly !

and MS Word, not MS Works, of course, gosh, my excuses !

UltraRecall is not the only outliner using an SQL infrastructure. In my view this has started to become quite common in information management software, even if the user interface looks similar to older versions that had other kinds of infrastructure. A case in point is ClipMate, the advanced clipboard manager. Another is Surfulater.

I think developers have good reason to go the SQL-way, particularly when trustworthy embeded solutions such as SQLite are available. No need to re-invent the wheel.

I use UR and other SQL-base software with no problem. The only thing I watch out for is making sure that a program is closed before backing up its database files. Contrary to documents and similar files, a database file may be backed up while in use, but this is not necessarily a good thing.

Alexander

Contrary to documents and similar files, a database file may be backed up while in use, but this is not necessarily a good thing

This issue is well documented. SQLite (used by URp, Firefox bookmarks, and may others) is NOT a multi-user database engine, so users must close the application before backing up. True multi-user DB engines (Jet, SQL Server, mySQL, etc) do not have this issue.