I've been evaluating the robust Windows knowledge bases (MyInfo, UltraRecall, IQ, MyBase) and MyBase is the one I keep paying more attention to. I like it so much that I'm now wondering what I'm missing because it gets very little regard from this forum and from the many review sites compared to the others. It's been around a long time so I find the lack of independent opinion about it within these circles concerning. Maybe it has improved a lot the last couple years after many had dismissed it? It's even on the three big OS's (Mac Windows, Linux) which makes me even more confused by the inattention.

Here's the features it has that I wish MyInfo would do:
-Nested tags/labels. This is a fairly rare feature across all outliners but MB does it well and has a really nice filter drill down when navigating via the tag/label tree.
-Two-way links (backlinks). These aren't 'inline'....in fact, does ANYONE do inline backlinks? Separate discussion. In MB they are part of the note/item metadata. This is made even more useful by a 'list all related items' search window that lists all the items that are directly linked to the selected item AND items that are not linked to THIS item but linked to the items it is linked to.

It ALSO does cloning/transclusion like the other mentioned solutions. I have yet to put a lot of content in it but the trifecta of backlinks, nested tagging, and item cloning brings together the best things about an outliner (structure and a sense of place) and a wiki (discovering and building upon unhierarchical info relationships).

It even now has a markdown option, which shows that it is 'currently developed' enough to be trendy (fairly big update April this year). As I am trying to learn the interface, I'm discovering a lot of customization like the ability to assign most functions to a hotkey (like inline time-stamps that I'm used to with OneNote). That feature alone helps get around a lot of interface clunkiness. It's GUI seems slightly less intuitive than UR/MI but more approachable than IQ.

Major features it lacks compared to the others:
-No custom metadata on items (like MI, UR and IQ do) which makes it more limited as a home for structured info like invoices and a book reference database.
-Can't open several windows of notes or pull 'floating windows' out of the main...uh...window. Though it can pop out a page in html so you can at least have a read-only side reference. I haven't experimented with running two instances of MB as a workaround (I do that with OneNote all the time).
-No 'map' view of pages and relationships like ConnectedText and IQ. Of course MI and UR don't do this either. Though MB will export a mindmap, which is neat but odd that there is no mind-map view within MB.

The only decent review I've seen on this is from the great Paul J Miller https://pauljmiller.wordpress.com/2013/06/20/144/ and his main complaint was lack of universal linking ..."The failure to be able to link things together is the most significant problem with this program, if this program did support universal links (supporting universal links into a database would automatically give it the ability to link to a note in a different database) then it would be the most awesome note taking program around and would blow away the competition."

I think Paul was reviewing version 6, it's now on 7.1 and has added the intra-database linking but made no progress on Universal Linking. And by intra-database linking mean dumb one-way hyperlinks. Which I'm pretty sure is par for the course of any other info manager when linking between databases. Though outside universal linking (where a hyperlink anywhere else in the OS opens the target note within MB) is no longer something I care about. I'm always switching between at least two PC's, so having to keep absolute paths between environments is a pain. Also my reference files are a combination of NAS, cloud, and local. Also my other referring and referred-to applications are a mix of installed/cloud and I CRIMP enough with those that I don't want to be tied down by needing hard links to always work between them on all my screens. My workaround is to create 'Universal Links' with Zettelkasten-like codes, which are simply unique text string identifiers that 'link' two or more things between disparate systems. Yes I understand that's not how Zettel systems work, I'm just borrowing the searchable-code-IS-the-link concept. All my codes start with Z, so my ONE unbreakable rule is if I see Z##### anywhere in my stuff, to DON'T CHANGE OR DELETE IT because it has a long lost brother somewhere buried in one of my piles of data hell. Pretty old school but I've been doing it for over a year now and it's surprisingly robust. Now I am free to swap out my task manager, re-arrange/move files, etc with any other component without worrying about breaking outgoing or incoming links because they are simply unique text strings. The extra step of searching for them instead of a simple click (and prayer that the link isn't yet broken) is well worth the freedom.

Paul's other concern was a major slowdown after the database went over 300mb. With version 7, MB probably fixed that by upgrading the database (like what Petko has apparently done with the MyInfo 7.0 beta). But I haven't done a 'load test' to find out.

I was concerned that I couldn't copy and paste content from OneNote into MB, as simple text or tables would end up as a graphic that I couldn't edit. The same content pasted into MyInfo/IQ just fine. But then I discovered MB's 'clipboard monitor' tool, which you turn on before doing an outside copy. Upon a detected 'copy', it opens a dialogue window asking how you want it to handle the content, and through that I was able to drop the OneNote content in just fine. So yeah it's not quite as smooth as MyInfo but it seems to offer more customization if you are wrestling with getting a webpage or something in there correctly. That's a recurring theme with MB. It's feature set is very large, with many extra plugins, but not always intuitive or smooth.

So it doesn't seem to be quite as good as UR and MI as being a desktop sidekick to other desktop programs. Particularly Outlook. I do a LOT of dragging and dropping Outlook messages into OneNote. Ok upon further quick testing neither MI or UR will do this either.... So to 'embed' email reference outside of OneNote means an intermediary save-as/export option from Outlook. Now I'm real temped to keep using OneNote as a project reference attachments Mule and just Universal Link to that page...

But, while not playing as nicely with others in real-time, MB does have a large array import/export options (like it will import Outlook stuff, which I don't care about since I see no value in a one-time snapshot from Outlook...unless I was 'archiving' a project I suppose). It seems to be heavily focused on file/directory management and might make a killer front-end to a reference directory. It even has 'duplicate file report' and 'compare folders' utilities in addition to the usual 'import file structure and create links' stuff that other applications commonly have.

It seems very stable. I'm running a 'portable version' on my work pc and haven't seen it stumble, whereas a portable MyInfo installation has been less consistent.

So at this point I have to wonder what I'm missing here. Why hasn't MyBase been mentioned as part of the age-old MyInfo vs UltraRecall discussion?

One reason, maybe: MyBase development proceeds at a glacial pace and the developer doesn't allow beta testers to talk about their experiences with the new features. The last update to version 6 was released nearly a year ago, and it's been a couple of years since the private beta for version 7 began. So there isn't a whole lot to discuss, and as a result the software (it's very good software!) might get overlooked when people are making comparisons.

Thank you for your in-depth review.

I was an early adopter of MyBase almost 20 years ago, but I didn't like it because -- at least back then -- it was a modular program. You bought the basic application, but then had to pay extra to get add ons. Not only that, I found the add ons didn't always work seamlessly. It doesn't look as if that is the case any more.

Steve Z.

Also the interface is distinctly unusual. The slowdown Paul refers to is still present, at least in my testing. And the hierarchical tags (if my memory is correct) are not what one would expect. For example

A
...1
....2
B
...3
...4

In CintaNotes, assigning tag 2 also assigns tag A. The result looks like this: A/2.

Selecting tag 2 in myBase assigns tag 2 only. To get the same result as CintaNotes, you must select the parent tag too. Again, this is based on my memory of myBase.

Jon

Are there no screenshots anywhere? It's a red flag in itself that the home page doesn't seem to have any.

cicerosc wrote:

Are there no screenshots anywhere? It's a red flag in itself that the
home page doesn't seem to have any.

Indeed, it's odd that the home page does't have any, but there are some on softpedia:

http://www.softpedia.com/get/Internet/Servers/Database-Utils/myBase-Desktop-Edition.shtml

(Click on "Screenshots" down on the lower left of the page.)

Jon Polish wrote:

Also the interface is distinctly unusual. The slowdown Paul refers to is
still present, at least in my testing. And the hierarchical tags (if my
memory is correct) are not what one would expect. For example

A
...1
....2
B
...3
...4

In CintaNotes, assigning tag 2 also assigns tag A. The result looks like
this: A/2.

Selecting tag 2 in myBase assigns tag 2 only. To get the same result as
CintaNotes, you must select the parent tag too. Again, this is based on
my memory of myBase.

Jon

Thank you for pointing this out. I did some checking and at first thought that your impression was no longer true because it shows the parent path of the tag both in the item metadata and the filtered tag list. However, if I tag with only the child, the item won't show when only selecting the parent tag. So you are absolutely right.

I'm not quite sure if I'd rather see it inherit parent tags as I'm trying to wrap my head around how to utilize nested tags AND cloning. Like I think I'd be more likely to use the cloning as an alternate classification hierarchy and treat the tags (MB calls them labels, in this discussion I consider tags/labels the same thing) more like descriptive attributes where the nesting has more to do with keeping a growing list orderly than to reflect a hard hierarchy.... nah I'm just BS-ing myself. I'd prefer they inherit and even transclude. Would be nice if this was an option.

I'll bet it's really annoying for Pierre to read these threads as they all tend to bend towards 'IQ can do all of these things'.

Ha, ha, ha! Very true (that IQ can do all these thing AND that Pierre will say this).

Jon

nathanb wrote:

I'll bet it's really annoying for Pierre to read these threads as they
all tend to bend towards 'IQ can do all of these things'.

Jon Polish wrote:

Ha, ha, ha! Very true (that IQ can do all these thing AND that Pierre will say this).

LOL, far from true, as IQ is Windows only (unless on a VM) , and doesn't really have nested tags yet

Pierre

LOL, far from true, as IQ is Windows only (unless on a VM) , and doesn't
really have nested tags yet

Pierre

I'm aware of Paul Miller's recent work-around to get IQ to act 'as-if' it does nested tags. That really says a lot IQ's core concept of 'everything is an item and any item can be a parent/child of any other item' and how ultimately flexible that is in practice.

MyBase has some nice features but I didn't really get along with the user interface and the fact that it became slow when it had lots of articles. I did try out the trial version of MyBase 7 but found that not a lot had changed on the user interface so I didn't make any great effort to put it to the test.

Right Note on the other hand has a very nice user interface but lacks many features which I use when they are available. Things like transclusion and clones, backlinks (what points here?), a hierarchical tagging system and universal links. The addition of these things would turn it into a true wiki.

InfoQube is nice, shame about the complex user interface but I suppose haveing all those features does complicate the user interface somewhat. At least IQ is configurable so you can strip out a lot of stuff you don't need and that simplifies things quite nicely.

Why can't there be an app like IQ with an interface similar to RN ?

Oh Well ......

Paul J. Miller wrote:

Why can't there be an app like IQ with an interface similar to RN ?

Most likely because RN is a straight "2-pane outliner" ***, whereas IQ has:

- Real, intrinsic, 1-pane outliner with columns, like OO or Ecco Pro
- Second pane for long documents
- Full featured Calendar
- Project management tools
- Scripting to automate your information management

That said, if you have suggestions to improve IQ's UI, I'm all ears !

***: 2-pane outliners aren't really outliners IMHO, they are a set of documents, arranged in a hierarchy

Pierre Paul Landry
IQ Designer
http://www.infoqube.biz

Pierre Paul Landry wrote:

Paul J. Miller wrote:

>Why can't there be an app like IQ with an interface similar to RN ?

Most likely because RN is a straight "2-pane outliner" ***, whereas IQ
has:

- Real, intrinsic, 1-pane outliner with columns, like OO or Ecco Pro
- Second pane for long documents
- Full featured Calendar
- Project management tools
- Scripting to automate your information management

That said, if you have suggestions to improve IQ's UI, I'm all ears !

***: 2-pane outliners aren't really outliners IMHO, they are a set of
documents, arranged in a hierarchy

Pierre Paul Landry
IQ Designer
http://www.infoqube.biz

Yes RN outlines are a strict hierarchy which means that it is a tree ( https://en.wikipedia.org/wiki/Tree_(graph_theory ) rather than a directed graph ( https://en.wikipedia.org/wiki/Directed_graph ) which is the structure in IQ. Directed graphs are better. However an outline may be included as a note type which is rather neat (it also has a spreadsheet as a note type).

With just a few more features RN would become much more useful, but those few features probably represent a hideous amount of programming work. I don't know.

Right Note is a lightweight which I find limiting but it does have an attractive user interface. InfoQube is more powerful and does much more but is also harder to use.

As for designing user interfaces, I know a good user interface when I see one but I am not very creative when it comes to original ideas for UI design. When things occur to me I will suggest them in the appropriate place.

nathanb wrote:

So at this point I have to wonder what I'm missing here. Why hasn't
MyBase been mentioned as part of the age-old MyInfo vs UltraRecall
discussion?

Maybe this is just a coincidence but maybe not

Having just taken a look around various note taking programs listed here I have noticed that the ones which are more successful and get more mentions in this forum and in other places are the ones which have forum in which users can discuss the merits of the program and interact with the developers.

Those which get little attention and are less successful are those which don't have a user forum (Right Note and MyBase).

There is also one where the forum used to be a lively interesting place but now the developer has largely abandoned the forum and it has become almost dead. That program is ConnectedText which seems to be on the decline.

That said, if you have suggestions to improve IQ's UI, I'm all ears !

MyBase, MyInfo, and UltraRecall all have something in common that makes them more approachable than IQ. Well two things. Obviously IQ is much more flexible and open-ended in it's workflow possibilities, and of course that makes it tougher to grasp. Assign from just complexity difference, IQ doesn't use ICONS to differentiate data types (as far as I can tell).

The vast majority of people (myself included) are used to interacting with software where the items convey strong sense of place and content. Imagine Windows explorer without folder icons indicating "a bucket of files or more folders" or files that were just a list of text like they used to be in the DOS days. Even logical links have well-established visual indicators of those little arrows on the icon and I think MB, MI, and UR all use those to indicate item clones.

Of course IQ is fundamentally different in that items don't have a place but show up when appropriate depending on the 'grid' being displayed and that attributes and grids are items too. So I can see why you wouldn't want to pidgeon-hole a subset of items to show an icon that indicates they are a thing, not just metadata about a thing (yes I sort of understand that the genious of IQ is that attributes can be actual full-featured documents too). So why pretend to have a 'sense-of-place' when the whole point is to have everything be a pivot on everything else?

I...don't know. All I know is that when I use the other three I instantly understand the overall scope and breadth of my notes and I commence to throw metadata and alternate clone hierarchies at them until the database starts to feel too complex to be manageable or clearly understood. I never mistake a tag for an item, or an item for an attribute, or a saved search (grid) for an item. I never 'lose sight' of my content which is almost always notes in the form of a document.

With IQ, I 'lose sight' of my notes almost instantly when I start to dink around with the attribute window which is a combination of simple attributes which may only 'describe' this particular item, categories common to hundreds of items, and saved-searches (grids) with little visual indicators to differentiate which is which until I click on one and experience a massive pivot away from the outline I was just getting the hang of. I really do understand how databases work and it doesn't take me long to wrap my mind around that I'm just looking at a different view and that my precious actual typed notes aren't any more buried than in the other info managers... It's just a level of abstraction beyond me feeling comfortable where my stuff is.

Is it possible to auto-generate a 'document' icon in the title of an item if html pane content exists? Therefore I can instantly differentiate between content and metadata within all contexts and never worry about 'losing sight' of my actual notes?

Along those lines, is it possible to auto-generate a 'grid' icon in the title for all named grids? In the 'properties' pane, the named grids are blue underlined text which is exactly how hyperlinks in many other contexts look. I've learned what they really are within this context, but I still have to remind myself to not think of those as hyperlinks. The properties pane shows a LOT of similar looking text that means very different things. On one hand the shear power of being able to jump to items and pivot to other views from that pane is glorious but on the other hand it adds to the mental overhead to have to mentally categorize the different text strings. Of course they are already categorized for me by the headings they are in and I do understand those. But it still adds a mental step. Again I refer back to windows explorer and what it would feel like to navigate without the icon column. Of course I'd be able to differentiate excel files and photos by the file extension and I'd know where I was at in the hard drive by reading the \path... but seeing that ubiquitous green excel icon under that folder icon saves me the mental deciphering step.

I hope you don't take this as actual criticism as I am able to walk through IQ and eventually understand what I'm looking at by stopping and thinking about it. When I do I'm always impressed at your achievement. I've obviously been conditioned to 'software for dummies' and my hangups are my fault, not yours. Just trying to convey how it feels to a normal 'power user' who is unable to make the mental leap to 'database geek' that IQ requires.

nathanb wrote:

So at this point I have to wonder what I'm missing here.

On the desktop environment (in my case PC one) three things in v.7:
1. Does not index PDFs.
2. WebCollector is browser specific, and there is no a *global* way to copy a web page or whatever else and paste it into MB without its addons (not so in UltraRecall and RightNote).
3. Linking files, drag'n'dropping, and moving items in the outline are not as "natural" as in other apps (example: an item is moved by a mix of drag'n'drop and menu selection).

Having said that I still use v.5.51. Max Ram used: 9Mb.

Anthony wrote:

nathanb wrote:
>So at this point I have to wonder what I'm missing here.
>
On the desktop environment (in my case PC one) three things in v.7:
1. Does not index PDFs.
2. WebCollector is browser specific, and there is no a *global* way to
copy a web page or whatever else and paste it into MB without its addons
(not so in UltraRecall and RightNote).
3. Linking files, drag'n'dropping, and moving items in the outline are
not as "natural" as in other apps (example: an item is moved by a mix of
drag'n'drop and menu selection).

Having said that I still use v.5.51. Max Ram used: 9Mb.

Thanks so much for this. Really helpful. This does reflect my impression and the main reason for this post. That it has all the features I tell myself I'm looking for but it feels like I'm talking myself into liking it more than I do. It's visual design is really nice, but as you've highlighted, it's a bit more clunky to use than others. Still haven't ruled it out though. I'm encouraged by the customizable keyboard shortcuts which could go a long way towards streamlining it.

Very first sentence of this thread: "I’ve been evaluating the robust Windows knowledge bases (MyInfo, UltraRecall, IQ, MyBase) and MyBase is the one I keep paying more attention to." - then: "Paul J. Miller's main complaint...links..."

Well, Mr. Miller has done some very valuable work re "outliners", shown on his blog; among other things, he tested them with "heavy load", and he said, re MyBase, that even with some 10,000 items or less, I don't remember very well, response times went sharply down (if I remember well, again), and so I would certainly not list MyBase among "the robust Windows knowledge bases".

As an aside, he discovered how to bold tree entries in UR, hence my question if that could be realized as some 1-key macro (and as toggle, ie with de-bolding the entry if it's already bolded, too); I never grasped UR's "user attributes" to begin with, so I unfortunately do not know where to begin with such a macro, but it undeniably would be a big enhancement to UR while almost certainly making any ephemeric interest in MyBase obsolete.

Since at the end of the day, 10,000 items is nothing, once you've got a conclusive way to get info into your bucket software, it'll be all about its organization then, and as said, I've had only minor complaints with almost 1 million of entries within UR.

(So the remaining posts/pages of this thread weren't of any further interest to me: be more concise, dear fellow software judges: Once you encounter some eliminator, don't pursue further, spare your time.)

22111 wrote:

(So the remaining posts/pages of this thread weren't of any further
interest to me: be more concise, dear fellow software judges: Once you
encounter some eliminator, don't pursue further, spare your time.)

Actually we had some good discussions going on after that first post and I'm a wiser CRIMPer for it! In my first post I acknowledge the indispensable Paul Miller's review but also mentioned that it was for an older version and that one of the most significant changes since that review was a significant upgrade to MyBase's underlying database. So it may be still true that MyBase slows down with growth, but I'm not sure that is a 'settled' issue at this point. However, though I've done infinitesimal testing with it, it does seem to be not designed for speed like Ultra Recall. I'm basing that solely on needing to manually 're-index' for a search to find any new entries for a 'current session' and that taking a few seconds for even a very small dataset. Even OneNote is very fast and non-intrusive about 'continuously indexing'. That doesn't bode well for '1 Million' entries in MyBase. I'm no programmer, but it 'feels' like MyBase doesn't entrench itself heavily within the OS like some of the others which I'm sure is why it seems to be the exact same animal on Windows, Linux, and MacOS. I'm leaning away towards using MyBase as a daily driver but think it is viable as a front-end for a large file archive that a person updates periodically.

Even though I find nested tagging and direct 'two-way' linking valuable, I'm now leaning towards Ultra Recall as my main system. The freedom of custom user meta-data for many different data types within the same database is very compelling. That means I can contain completely different kinds of Knowledge Bases under different branches of the main hierarchy but still 'cross link' individual items between those different data sets when needed. Sure I'd rather have explicit 'backlinks' and nested tagging, but the cloning (and quickness that UR lets you clone things) enables me to emulate those two things reasonably closely, especially considering the custom attributes when necessary. I'm a heavy Outlook user at work and am only just realizing how UR can level up many aspects of that flow.

When you say you have '1 Million' UR entries, is that within a single database? How big is that file? I think the UR devs say the database limit is 2TB but I'm curious to know how big the larger ones in the wild are.

If it helps, I just tested version 7 with my test database of 16885 notes that was originally made with version 6. The database is marginally faster, BUT I am using a newer computer than my original testing. A fairer comparison would be to install version 6 on my current setup, but I no longer have it. My guess is there will be no difference.

Jon

Jon Polish wrote:

If it helps, I just tested version 7 with my test database of 16885
notes that was originally made with version 6. The database is
marginally faster, BUT I am using a newer computer than my original
testing. A fairer comparison would be to install version 6 on my current
setup, but I no longer have it. My guess is there will be no difference.

Jon

That is helpful. Thanks Jon. This exercise has reminded me how important speed is with something like this. I'm going to try and see if I can get Ultra Recall to emulate 'backlinks' and nested tags with just cloning. Maybe it's better to push the limits of a few features done well than to juggle more bells and whistles.