- User Since
- Jul 23 2018, 3:22 PM (20 w, 1 d)
- Code cosmetic
- Add singleShot stuff
everything could have happened, e.g. the search bar got deleted because the view got delete
I would also advise against calling processEvents()
It is something that comes up every once a while, but has just proven to be a bad idea over the last ten-something years.
Hmpf! Some googling didn't help. Just some thoughts.
That always leads to evil things, like e.g. what happens if you press the X button of the view/window during that.
- Check every 500 lines or 100 machtes, whatever is first
- Fix false check due to wrong math as long no match was found
I would probably not make the button that big
Sun, Dec 9
Not well tested but looks so far pretty promising.
Um, I'm bit confused when I now look at the diff 1/2 here???
- Add @since hint to speller.h
- Add some blank lines to speller.h to be looking a little bit nicer
I used this patch now for a while and most of the time it works (for me) charming. But there is some odd behaviour when you try to do some "no normal build" stuff. Then is these auto-close feature a little bit annoying.
So any tip is appreciated. Will look at it some time later.
Yes, there is now a limit for how many highlightings are done.
If its too much, we skip that, else the rendering will break everthing down ;=)
In this state was this patch only a simple attempt to avoid the worst case.
With your new improvements is it much less likely to happen, but still possible. I do not like to insist of this patch it's only an offer, which could be improved.
This needs now some user testing ;=)
I have tried this one. Yes, works much more better!
Nice to see that my own (unloved) patch wake your interest to dig into that issue :-)
Sat, Dec 8
Would be good you take the time and do that test @cullmann When you then think it is still useful to soften that limit let me know how.
To become that code more smart as you suggest I have right now not the mood, sorry.
I think this patch is good enough to avaoid the biggest trouble.
- The mentioned crash could I not reproduce, but after 5 million hits and a felt eternity I had killed Kate
- There is an issue now with the wrap message, where is now the same close button. Any ideas what's wrong?
- I like to suggest to get rid of most of that S&R code and use instead code from the Search plugin. There is the S&R done in an own thread and did not block the UI.
That line is a little bit long. Let me know if I should wrap them in shorter pieces.
Thanks for your input Dominik.
In the meanwhile I noticed (again) these "Smart-Return" function (I noticed, and used that a couple of years for a short time period, but forgot that somehow) A reverse version of that should be nice. And somehow the possibility to use that automatically without to todo some special key-stroke when join lines by simple remove the newline character by Del-Key.
@cullmann I thing you could give a short hint
Fri, Dec 7
Frameworks Ping! (?)
Please addBUG: 389181
To its own line in the Summary section. See https://community.kde.org/Infrastructure/Phabricator#Formatting_your_patch
Thu, Dec 6
Without any special "Oh, that would be nice", I will stop now with further S&R
Wed, Dec 5
@cullmann Did you notice this? The other patches did you apply very quick, so I hope I didn't unsettle you with my comments.
Just a hint/question from someone who didn't try this particular patch.
Tue, Dec 4
I like the approach to move much into a 2nd level menu because I'm easy annoyed by long menus where I have to scan and move a long way down.
Mon, Dec 3
- Updated as accepted
- Cleared Test Plan to avoid unneded stuff in commit message
Sun, Dec 2
I think the text could be "Cannot run command: %1\nWork path %2 does not exist."
No sure if that's a request to me to change that, or only a thought for further discusion.
My taste is not to include the argument into the sentence, so I offer a slightly modification. Nevertheless looks that to me a little long.
The text (the look) may OK but I think could be improved. Please give exact advice.
Sat, Dec 1
It looked strange on the list of names of all committers.
Fri, Nov 30
..but I cannot remember we didn't ask contributors for their real name for contributions.
Well, neither of that. My "online me" is loh.tar but that was rejected by Phabricator, so the dot had to go. Lastly used ngraham uppercase letters, what looks even more strange to me :-)
May I ask how does this here will progress? You seemed to request a change which was not clear to me, but it's anyway "green flagged".
Only 'm_view->' => 'view()' has plenty of hits, the others only a few.
Thu, Nov 29
I can split it if you like in two diff
- 'm_doc->' => 'doc()->' ~100 hits, perhaps not the best idea? A search for doc() has 24 hits (without this patch)
- 'm_viewInternal->m_cursor' => 'cursorPosition()' ~10 hits
- Prefer function calls for member excess
- Merge KateViewInternal::updateView with KateViewInternal::doUpdateView
- Add documentation to related functions and some near by cosmetic. Well, "Better no docu than bad docu", so if that is to difficult to check now, I can remove it and ship as own diff
- Complete commit message
- Remove Testplan, I'm always surpsised to see that later in the commit message
Wed, Nov 28
Bug302689 - Impossible to delete unwanted dictionaries
No, but that's fine. One thing less where I can make mistakes :-)
Should your next question be regarding author ship, take a look at edbf2e36c805
Um, no(?) In that case I would have named this patch "Add stretch to ensure Dump button is expanding"
Sorry, next time I try to describe the pics more precise.
I think that's all.
Well, it looks less beneficial as previous anticipated by me, but I think it's still the right direction.
we can have more improvements...in a new patch.
Tue, Nov 27
Hi, I accidentally went by.
How about to add an icon to some (main) toolbar, if any is present?
That would the need to show such a message reduce to cases where no tool bar is available. If that icon was added, and not already there, could that pop up a small tool tip to get some attention.
A quick search here in my browser give no hit for "isVisible". There should be no need to show such message when the action is somewhere seen.
- Remove comments as requested
- Change Button text and icons as requested
- Avoid gab of Close button to the bottom when window is smaller
Mon, Nov 26
Still one should take the additional input into consideration ;=)
To me, "Duplicate..." sounds more natural, and I don't think I've seen "Copy As..." in English in very many user interfaces.
@ngraham Thanks for the VDG/screenshot advice, will hopefully remember that the next time :-)
Sun, Nov 25
I am not sure if the normal join lines should try to be that clever.
Glad to read that :-)
Sat, Nov 24
I would like to avoid these Arcanist. But every hint what I have to be aware of is much apricated.
Fri, Nov 23
This works so far in most cases I have tried.
Thu, Nov 22
It would be nice someone could explain, or point me where to read, why there is no (more) direct use of Sonnet stuff in KTextEditor/Kate.
Tue, Nov 20
@davidedmundson Wow, that was fast :-)
Mon, Nov 19
Um, really 9999? I have used now 9
It just use big enough number to not have "Context not available."
The added stuff at the ConfigWidget looks to me not perfect but OK. Suggestions are welcome. Some thoughts/questions
Sun, Nov 18
- Remove preprocessor protection of QDBus availability
- Increase kpartgui version to 84
Oops! - Fix to give anonymous session a name
- Avoid own entry in window/task-manager
Fri, Nov 16
Because this a little extensive patch, I guess you will need (and should take) some time for a review.
Nov 5 2018
Nov 3 2018
Oct 22 2018
Oct 20 2018
Aug 13 2018
I would still would want to integrate get your Normal parser as a Markdown parser if you are OK with that.