Erasure of palette search string in 3.3 is much more inefficient (time-consuming) than 3.2.3

• Nov 23, 2019 - 02:13
Reported version
S4 - Minor

Of course it can be claimed that this is not-important, as it's only about 1.5 seconds of "frozen", but those who use palette searching quickly (self included) will more than likely find this a bother, even if it is only a few seconds, especially since 3.2.3 had no such delay.

To be more precise: deleting the text within palette searching is not snappy enough as it was in 3.2.3, and not only as in backspacing the text, but also with a quick CTRL+A and then backspace to clear the textbox. This was practically instantaneous in 3.2.3.

Here is a quick demonstration:


There is no delay from highlighting the text in the search box via [ctrl+a] and pressing [backspace] on a keyboard, so the time that it shows the search-text highlighted is all some sort of inefficient "calculation" in 3.3.2.

Although I wouldn't call this a major bug, it's not merely just a "suggestion", as it ought to be looked into when time allows as to what is happening to warrant such a 'frozen' moment for the user.


In reply to by dmitrio95

Thanks for tip. Even better for me, I utilize an autokey script to [search palette command]+[ctrl+a]+[backspace] which worked wonders in 3.2. Maybe i'll replace the last part with [ctrl+backspace].

Anyone else recognizing slowdowns? If I'm the only one, I suppose I'll have to "check things out." Then again, it wouldn't make sense that it works quick in an earlier version but not now, yet end up being the only one with the issue.

In 3.2, I was able to [palette search]+[clear the text-field]+[type search key]+[down+enter] without any slow-down present in the UI at all. If 3.3.x could be just as snappy over here, it would be really nice, especially with the new [apply current palette element] command.

In reply to by worldwideweary

Maybe I should request users in the forums to contribute if they're willing to do quick screen captures of erasing a few examples of a search-key or two in the palette to see if this is an issue with other systems, or to provide proofs of no slowdowns, and if some have the issue maybe it'll be looked into by the people who changed the code from whatever 3.2.3 was doing to 3.3 to see if there can be found some inefficient code.

And also for what it's worth, here's a screencast of doing the same action in 3.2.3: no delay at all on the same system: