InfoQube IM - Community

Submitted by WayneK on 2018/01/01 12:17

What's the easiest way to ensure that the item sequence doesn't get accidentally scrambled? Worst-case situation: you have a long sequence of subitems that you inadvertently sort alphabetically. The result would be a time-consuming mess to untangle.

I suppose you could get a roughly accurate re-arrangement by sorting by modified date, but this won't work for items that were edited or inserted later.

When I take book notes, I use a sequence column so that every item is assigned a page number and sequence number on that page. So there, I don't have to worry.

In other cases, there are no page numbers to assign (eg web notes), and I've fallen into the habit of taking long sections of notes with no sequence assignment. I started getting nervous about this so I decided to add a sequence column for these notes and in the process of doing this I partially scrambled my notes and had to re-read the article to fix it (ironic).

It would be nice to have a failsafe system for preventing this problem that doesn't rely on manually inserting sequence numbers (though the "re-number" command does make this pretty easy).

Maybe there's something in outlining labels that would accomplish what I want (never used them). But an outline is still subject to sorting and re-labeling, I presume. Or maybe a combination of item created/modified sorts is good enough.

Any ideas on this would be appreciated. I don't need a detailed description but just a point in the right direction.

Wayne

Comments

[quote=WayneK]

Worst-case situation: you have a long sequence of subitems that you inadvertently sort alphabetically.

(...)

It would be nice to have a failsafe system for preventing this problem that doesn't rely on manually inserting sequence numbers (though the "re-number" command does make this pretty easy).

[/quote]

>you inadvertently sort alphabetically

In practice, how does this happen ?
Would a confirmation dialog help ?
Or else would implementing Undo on sorting commands help ?
Finally, would you like the sort commands totally disabled in some situations or as a setting for a grid

Pierre_Admin

IQ Designer

It happened to me today and I still don't know exactly what happened. I clicked on the wrong tool icon and somehow one of my items went to the top of the sort. I had to compare the source to my notes line by to make sure everything was in the correct order. Not something that happens regularly but could be a pain if you have a long string of notes.

I'm not going to try to create exact scenarios but with all the sorting and filtering options it doesn't seem farfetched to me that you could inadvertently scramble your items and not be able to restore them to the manual sort you created. I've certainly applied commands to the wrong items because I wasn't paying close enough attention to what was selected.

I don't think it's worthy of a program change. I was asking if there's some way to accomplish what I'm doing right now without having to manually insert sequence numbers. If you were going to do something, I guess I'd favor the undo.

Wayne

I'm not particularly concerned about losing the ordering of TLIs in a grid, but I am concerned about losing my ordering of subitems under a parent item. So I think it would be great if the manual ordering of subitems could be locked in some way, and protected against anything that would change the ordering until the lock is intentionally unlocked. So maybe instead of being implemented at the grid level as you suggest, it could be implemented at the item level.

Jim,

Yes, I agree. TLI's by nature tend to be stand-alone topics whose order is less important. For myself, subitems are often notes taken on a research source, and the order is paramount.

I just returned to a TLI and found that the subitem order has been reversed. Wasn't an issue because it was only two subitems but manual sort is turned on and I know I didn't reverse the order myself.

I haven't tried to nail down when this occurs but it's happened to me quite a number of times. I have an impression that it's related to hoisting and applying column filters, which I use a lot.

Wayne

One time I used subitems to record a sequence of events in chronological order, one subitem per event. I didn't use another column to hold numbers to preserve the order. Don't know how it happened, but somehow the subitems got rearranged, and I lost all sense of the sequence, and I couldn't really reconstruct it. Since then I've been gun shy about using subitems to make such a list. There are workarounds that can protect the ordering, of course, but it would be nice to have an easier and sure method.

Posting to confirm what I feared: it's quite common for subitem sorts to get scrambled even if you have manual sort on. I'm going through old notes right now and approximately 50% of my subitem notes have been re-sorted alphabetically. If I didn't have a sequence column set up, it would have been a real problem. Even with the sequence column it's a pain because there are so many of them to re-do.

I certainly didn't sort these manually. Again, my suspicion is that some combination of hoisting and column filtering causes the unwanted subitem alphabetical sorts.

Wayne

Hi Wayne,

Can you reproduce the problem and provide steps + screenshots of your grid + columns?

Like you, I work with a lot of items and with all kinds of sorts* and this problem isn't happening. E.g. among many other things, I wrote a thesis in IQ and sorting was obviously important; both numbering -- number column for column sorting -- and manual sort were used, and it never failed me. Stuff can happen though, and I might have been lucky.

Thanks

* Column sorting ("sort bar"), column/field grouping ("group by bar"), source bar sorting ("source item sort"), manual sorting ("None"). Plus the other subtleties... like when there are no fields in the column sorting mode OR when "sort items **and sub items**" is off, and when the sort subitems descending" or "ascending" function is used. Those last ones can be tricky as this is where users sometimes destroy their "manual sort". [Edit : not to mention that each hierarchical display mode is saved with its own sort options -- one needs to be vigilant.]

- To prevent that, maybe sorting (manual) should be recorded only when when the "None" option is selected, and in no other case (even in the cases where sort sub items isn't selected)? But this has its drawbacks too.

- OR maybe with a special function that would save the sort only when the user saves it explicitly...? (And that has its drawbacks too : what if you wrote 20 pages worth of text and forgot to save, and then resorted the whole thing with "sort subitems ascending/descending"?).

For me, the best failsafe has been to renumber items periodically, which is pretty much like a "save sort" while keeping the automatic manual sort saving working in the background.

After some tinkering, I was able to reproduce the problem. I never noticed it before, probably because I rely on numbering a lot. Anyway, thanks Wayne.

Steps

1- Create a manual sorting. Refresh to make sure it's recorded.

2- Switch to an automatic sort : Sort with source bar or sort bar.

3- Manually move an item while in the automatic mode.

4- Revert to the manual mode : the original manual sort is gone and replaced by the previous mix of automatic + manual.

I think manual sorting should never be recorded, except while in manual mode. OR probably better (as to record a mix of manual + automatic can be useful) : like Pierre suggested, if items were manually moved while in an "automatic" sort mode, a dialog should warn the user when switching back to manual mode : "record the new sort or revert to the last manual sort? " or something like that. Same when "sort subitems" is turned on/off.

Thanks, Armando. No, I can't reproduce it. I generally don't see it happen. I return to notes I'd worked on earlier and find the order scrambled. Your method of re-creating the problem isn't what's happening with me, I don't believe.

If I can figure out what triggers it in my case, I'll post it.

Wayne

Hi Wayne,

See my last message where I was able to reproduce it. Tell me if you can too.

I wasn't able to reproduce it consistently:

1) First I tried 6 items using your steps. When I returned to manual sort at the end, it correctly put things back in their original order.

2) When I increased the number of items to 10, it failed to return them to their original order at the end.

I'll spend some more time on it later.

Wayne

Ok, thanks for the feedback! I'll look at it again later too.

If I create a column to hold numbers to preserve a manual ordering of numbers, and use Edit > Renumber Items to fill it - only one sequence of numbers is created for the selected items, whether they items are TLIs or subitems. See screen capture below:

Would it be useful if Edit > Renumber Items created a separate numbering sequence for each group of sibling subitems?

I don't thing sorting would work if the sorting column had 1.1.3 as you'd end up with 10-19 being before 2:

1
1.1
1.10
1.11
1.11.1
1.2
1.3

Pierre_Admin

IQ Designer

[quote=jimspoon]

Would it be useful if Edit > Renumber Items created a separate numbering sequence for each group of sibling subitems?

[/quote]

(It might not be pretty, but this is how I do it nevertheless. It works well as a workaround. For the indentation an the actual outline numbering, there's the outline styles, etc.)

Well, my idea was to have the "renumber items" function work recursively with items/subitems, so that each level would have separate numbering. And the numbering would be 1, 2, 3 for each, not using the 1.1.1 kind of numbering. But that doesn't seem to be the best solution either, especially considering that a subitem can have multiple parents.

I don't know how the ordering of subitems under an item is stored internally in IQ.

But for a method that would be visible and "saveable" by users, how about a field for the parent item that would contain a comma-separated list of the ItemID numbers for that parent's child items? Then there could be a "restore subitem order" function for such a field.