LibreOffice Gets Flicker-Free OpenGL Transitions

[Mavman]

O que que o cu tem a ver com as calças?

Epa, Tem graça, Tava a interrogar-me sobre o mesmo?!?! Ahahah... Não vejo relação...

Then again... This made me remember that good old scene from MIB II where those funny little aliens that live in a locker all sing a chant in honor to gnome... A-ha... I mean, to agent J

[man-walking]

A bit offtopic: does anyone know why LO Windows version since 5.3 started to show a sluggish interface?
For ex. menus draws pathetically slow on any pc and Windows version (7, 8... 10)

[unixfan2001]

GTK3 and inspired by GNOME HIG. That's how you build a Standard Desktop

Firefox is doing the same

'cept Firefox fucking doesn't!

GTK3 is only used as a backend. The frontend is rendered through a custom toolset.

[cj.wijtmans]

Hmm no vulkan and multihreading? How about fixing the damn loading of spreaedsheetes and searching properly.

[DrYak]

Next breaking news: there is no major company paying to get Qt usage in Linux.

Nokia used to be one.
(The Elop and the whole Microsoft fiasco happened to them)

Now it's just a bunch of small companies :
- Canonical used to push Qt as part of Unity.
- Jolla is pushing Qt on GNU/LInux smartphone OS, becaue they are basically the continuation of Nokia's Meego.

[duby229]

I couldn't care less even if they ripped out all animations from LO.

What I need them to do is to introduce multithreading into LO. It's currently so unbearably slow with image-heavy documents or spreadsheets with more than 100,000 rows of data and multiple worksheets that even Google Docs or Microsoft Office Online are countless times much faster and more pleasant to use at this point of time.

And I'm on an i7 Skylake processor.

I just don't get why people would use a calculator that way. Because that's what spreadsheet really are, fantastic calculators. You most certainly, beyond any doubt, are using spreadsheets wrong.

[schmidtbag]

I just don't get why people would use a calculator that way. Because that's what spreadsheet really are, fantastic calculators. You most certainly, beyond any doubt, are using spreadsheets wrong.

I strongly disagree. You do realize spreadsheets existed on paper form, before they were computerized, right? Yes, programs like Calc and Excel are fantastic at doing calculations but are much, much more than just calculators. However, what I would like to clarify (and what you may have had in mind) is spreadsheets are not databases, and should not be used as such. That's something people really seem to get wrong pretty often, in which case you would be right.

Regardless of whatever you want to classify them as, that has absolutely nothing to do with the fact that Calc needs to be more multi-threaded, and unlike most programs, this is one that could be done relatively easily. Just spawn a new thread for each selected column and/or each selected row. For most equations, it doesn't matter when the thread finishes. Right now, I'm working on a spreadsheet with about 80k rows (that's just how the data was given to me) and a simple search-and-replace can take as much as 30 minutes. Sure, I'm using a Core2 Quad, but the fact of the matter is only 1 core is being utilized. Meanwhile, every other Calc process is locked up until this task completes. This is a very poor design.
EDIT:
Believe it or not, but most of the wait time has to do with Calc highlighting the cells that were replaced. Calc is weirdly and horribly slow when it comes to selecting groups of individual cells rather than entire rows or columns. In other words, this is a performance problem that has nothing to do with calculations.

[duby229]

I strongly disagree. You do realize spreadsheets existed on paper form, before they were computerized, right? Yes, programs like Calc and Excel are fantastic at doing calculations but are much, much more than just calculators. However, what I would like to clarify (and what you may have had in mind) is spreadsheets are not databases, and should not be used as such. That's something people really seem to get wrong pretty often, in which case you would be right.

Regardless of whatever you want to classify them as, that has absolutely nothing to do with the fact that Calc needs to be more multi-threaded, and unlike most programs, this is one that could be done relatively easily. Just spawn a new thread for each selected column and/or each selected row. For most equations, it doesn't matter when the thread finishes. Right now, I'm working on a spreadsheet with about 80k rows (that's just how the data was given to me) and a simple search-and-replace can take as much as 30 minutes. Sure, I'm using a Core2 Quad, but the fact of the matter is only 1 core is being utilized. Meanwhile, every other Calc process is locked up until this task completes. This is a very poor design.
EDIT:
Believe it or not, but most of the wait time has to do with Calc highlighting the cells that were replaced. Calc is weirdly and horribly slow when it comes to selecting groups of individual cells rather than entire rows or columns. In other words, this is a performance problem that has nothing to do with calculations.

I'm no master spreadsheet designer, but I've made enough that I understand there is a maximum complexity that once reached a human being just can't comprehend anymore. And it's because of the way spreadsheets are supposed to be organized.

I'm not in any way arguing about lack of multithreading. I'm just arguing that giant spreadsheets that can't be understood is his particular problem. And it's probably your particular problem too from your description. The only person in this world that can look at your spreadsheet and understand it is probably you alone.

[Marc Driftmeyer]

I just don't get why people would use a calculator that way. Because that's what spreadsheet really are, fantastic calculators. You most certainly, beyond any doubt, are using spreadsheets wrong.

You clearly have never done much scientific data capture of live recordings. Take an Axial Flow Fan. Place sensors along the channel to map out it's Velocity Curve and determine how close theoretical to actual is matching in say, Fluid Mechanics Theory. Writing a database to do such work for long durations is something done last.

[MoonMoon]

Yeah, they failed. All of them failed. That's the point!

The time you used to troll around in this thread could have been much better spent fixing bugs in Gnome. Oh, wait, I forgot that you are just a troll, not a contributor.