[Ann] CintaNotes 2.5 Beta 3

[gunars]

For comparison, the Remarks field (and Title and Text) looks like it works ok.

Title, Text and Remarks field use FTS search (MATCH operator) - hence no problem. (But only when SIW=off).

So to summarize, MATCH is used for the Title/Text/Remarks fields if SIW=0. For all other combinations of field and siw, LIKE is used. The MATCH properly differentiates between accented and plain chars, while LIKE does not. Is that right?

Alex, I think that's fine for now. The first priority was to be able to enter the accented characters properly and that's working very well now. For searching, the examples I've used are just test cases. In practice, I would never be searching for just an individual character like that. When used as part of a normal word search, I think it will work just fine in 99% of the cases. A comment in the help is enough to explain why it is the way it is. You can then suggest using just Text and Title with SIW=0 if this is a problem.

2) I've also noticed that entering a char, even a plain English one, converts from upper to lowercase (I can actually see it make the swap, so I guess I'm not as old/slow as I think ). But this happens only for the first char, so it looks a bit odd ('DDD' --> 'dDD'). Also, I've had a few times when an uppercase letter did not get converted to lowercase, I think more often if the right Shift was used (this has nothing to do with accented letters). This does not happen consistently though, but I haven't pinned down the conditions yet.

Unfortunately I'm unable to reproduce this issue. If you can sometimes reproduce it, perhaps you could record a log file? This would help to track this down.

If focus is already in the search box, typing 'DDDD' puts 'DDDD' in the search box. If focus is not in the search box, typing 'DDDD' moves the focus, puts 'DDDD' in the search box and then pops the first letter to lowercase 'dDDD'. This is a minor cosmetic issue. I mention it only in case there is some side effect.

Gunars

[CintaNotes Developer]

My setting is "medium". Nevertheless, i can live with this. It's just, that my expectation was different. I expected it to behave like a click on a scroll-bar. The control for the tag sidebar is a different one, though. No problem.

AFAIK the standard scrollbar doesn't handle double-clicks at all, so I don't quite understand what was the expected behavior

[Thomas Lohrum]

If focus is already in the search box, typing 'DDDD' puts 'DDDD' in the search box. If focus is not in the search box, typing 'DDDD' moves the focus, puts 'DDDD' in the search box and then pops the first letter to lowercase 'dDDD'. This is a minor cosmetic issue. I mention it only in case there is some side effect.

Hi Gunars,

i can not reproduce the issue, though your steps are precise. Maybe you can provide your settings file. If so, you should send to Alex directly, as uploading files in the forum does not work.

Thomas

[CintaNotes Developer]

Depending on CN's implementation it might be possible to set the collation or maybe write a user function. I know too little (actually nothing) about CN's implementation, so it is best to trust to Alex's skills here.

CN uses a custom open-source Unicode library for SQLite which approximates the correct Unicode collation, but
is much simpler and lighter than a full-blown solution like the ICU library (which is 18 MB!)

One workaround might be to let the user know about the problem. For example, when an accented character gets entered, let the user know, it can not be used for search. Respectively show the message only depending on the SIW setting. Also a hint to turn off SIW to support the search.

Since our resources are very limited (and this will stay that way while we are selling less than 100 PRO licenses per month), I tend to be very choosy about what features and problems need to be attended to. I consider the votes on the roadmap to be a good enough indicator of user demand. So let's wait till this issue collects some votes. Maybe this problem is not a real problem, like Gunars said, and our time would be better spent implementing Undo or some other feature from Top 10

[CintaNotes Developer]

The one that's currently available is listed as requiring iOS7. The same for the Evernote out there now, although I think a previous version did work with iOS6.

Interesting, I didn't know that. Thanks for the info.

[CintaNotes Developer]

So to summarize, MATCH is used for the Title/Text/Remarks fields if SIW=0. For all other combinations of field and siw, LIKE is used. The MATCH properly differentiates between accented and plain chars, while LIKE does not. Is that right?

That is nearly correct. For most cases, this explanation would be enough.

In reality though it's a bit more complex: if SIW=0, but the query contains any punctuation characters (which are not contained in the full-text index), then CN uses 3rd approach: custom function LIKEWORD, which is a custom LIKE implementation which only matches starting from word boundary (SIW=0, remember?).

Alex, I think that's fine for now. The first priority was to be able to enter the accented characters properly and that's working very well now. For searching, the examples I've used are just test cases. In practice, I would never be searching for just an individual character like that. When used as part of a normal word search, I think it will work just fine in 99% of the cases. A comment in the help is enough to explain why it is the way it is. You can then suggest using just Text and Title with SIW=0 if this is a problem.

Good to hear that. Let's also see how many votes will the roadmap issue http://roadmap.cintanotes.com/topic/354 ... g-results/ collect.

If focus is already in the search box, typing 'DDDD' puts 'DDDD' in the search box. If focus is not in the search box, typing 'DDDD' moves the focus, puts 'DDDD' in the search box and then pops the first letter to lowercase 'dDDD'. This is a minor cosmetic issue. I mention it only in case there is some side effect.

Doesn't happen in my case, sorry. This is why having the log wouldn't hurt

[gunars]

If focus is already in the search box, typing 'DDDD' puts 'DDDD' in the search box. If focus is not in the search box, typing 'DDDD' moves the focus, puts 'DDDD' in the search box and then pops the first letter to lowercase 'dDDD'. This is a minor cosmetic issue. I mention it only in case there is some side effect.

Doesn't happen in my case, sorry. This is why having the log wouldn't hurt

OK, I sent the log, db, setting and screenshots to support@cintanotes.com. It's strange, since I have no trouble replicating it. I'll try it on another machine later.

By the way, for all of my testing, I've been using the portable version.

[gunars]

If focus is already in the search box, typing 'DDDD' puts 'DDDD' in the search box. If focus is not in the search box, typing 'DDDD' moves the focus, puts 'DDDD' in the search box and then pops the first letter to lowercase 'dDDD'. This is a minor cosmetic issue. I mention it only in case there is some side effect.

Hi Gunars,

i can not reproduce the issue, though your steps are precise. Maybe you can provide your settings file. If so, you should send to Alex directly, as uploading files in the forum does not work.

Thomas

Hi Thomas,

Alex tracked it down with the info in the logfiles. The behaviour I was seeing was only if Option->Tags was set to Lower Case Only. Otherwise it was fine. A minor issue, but I brought it up since the release notes mentioned some work had been done (triggering search with uppercase letters) in this area for beta 3.

Gunars

[gunars]

Alex, congratulations on the release of 2.5!!!

Gunars

[CintaNotes Developer]

Alex, congratulations on the release of 2.5!!!

Thanks, Gunars!! As you can see I've fixed the "dDDD" problem;)