Slashdot Mirror

I tried the previous version via mythbuntu, and found it so inconsistent and ugly (all those screenshots are from a single theme) that I even filed a bug report about it, but it was marked wontfix on the basis that a new version with new theme would be here in 6 months -- looking at the screenshots on the website I don't see much sign of improvement, can anyone who's used it comment?

Your screenshot is from three versions ago, 24+ months. Myth has received a full UI rewrite since then. I also don't know what version of the myth website you are referring to, but it displays several of the new themes in screenshots.

http://www.fecitfacta.com/Arclight/Welcome.html

As an example of the new UI.

I tried the previous version via mythbuntu, and found it so inconsistent and ugly (all those screenshots are from a single theme) that I even filed a bug report about it, but it was marked wontfix on the basis that a new version with new theme would be here in 6 months -- looking at the screenshots on the website I don't see much sign of improvement, can anyone who's used it comment?

obvious question: workflow / parameters?

Generally I use the foreground select tool* to select the smallest area to cover the object, then grow selection by a few pixels so that none of the object's edges are poking out and confusing it, then filters - map - resynthesize (ie, use the plugin rather than the script), and have the "tilable" options disabled since they tend to grab samples from the opposite edge of the image (if I want a tilable image, I'll use the tiling filter separately...)

Probably the biggest factor for simple success is to have the object you want to remove be on its own (surrounded on all sides by similar textures) -- if it isn't, then you need to do things the long way -- eg, if you want to remove the leftmost wheelchair from this image, and you want it to be replaced by grass when three of its borders are touching non-grass, then you'll find that it ends up somewhat messy since it attempts to merge four different edge textures. In this case you'll need to copy a section of your desired fill texture (ie, a rectangle of pure grass) into a separate image (specifically, a single layer image with no transparency); then on the original image select the object to remove, open resynthesizer, and select the "fill texture" image as the texture source; this way the generated texture will both match the surroundings of the original as much as possible, while being filled with the "surroundings" that you've specifically chosen. Having taken a sample of "pure grass" and a sample of "pure stone", then removing the top and bottom halves of the wheelchair with each respectively, the results are nicer. (with the exception that the first two images were produced with a mouse and twenty minutes of careful selecting, and the final one was 5 minutes work with a laptop nipple, so there are still some bits of wheelchair poking out of the sides...)

Incidentally, does photoshop have SIOX yet? Having the features "vaguely scribble in the general area of an object to have the object selected precisely" and "automatically and realistically remove a selection" could potentially combine to form "one-click realistic object removal" \o/

obvious question: workflow / parameters?

Generally I use the foreground select tool* to select the smallest area to cover the object, then grow selection by a few pixels so that none of the object's edges are poking out and confusing it, then filters - map - resynthesize (ie, use the plugin rather than the script), and have the "tilable" options disabled since they tend to grab samples from the opposite edge of the image (if I want a tilable image, I'll use the tiling filter separately...)

Probably the biggest factor for simple success is to have the object you want to remove be on its own (surrounded on all sides by similar textures) -- if it isn't, then you need to do things the long way -- eg, if you want to remove the leftmost wheelchair from this image, and you want it to be replaced by grass when three of its borders are touching non-grass, then you'll find that it ends up somewhat messy since it attempts to merge four different edge textures. In this case you'll need to copy a section of your desired fill texture (ie, a rectangle of pure grass) into a separate image (specifically, a single layer image with no transparency); then on the original image select the object to remove, open resynthesizer, and select the "fill texture" image as the texture source; this way the generated texture will both match the surroundings of the original as much as possible, while being filled with the "surroundings" that you've specifically chosen. Having taken a sample of "pure grass" and a sample of "pure stone", then removing the top and bottom halves of the wheelchair with each respectively, the results are nicer. (with the exception that the first two images were produced with a mouse and twenty minutes of careful selecting, and the final one was 5 minutes work with a laptop nipple, so there are still some bits of wheelchair poking out of the sides...)

Incidentally, does photoshop have SIOX yet? Having the features "vaguely scribble in the general area of an object to have the object selected precisely" and "automatically and realistically remove a selection" could potentially combine to form "one-click realistic object removal" \o/

obvious question: workflow / parameters?

Generally I use the foreground select tool* to select the smallest area to cover the object, then grow selection by a few pixels so that none of the object's edges are poking out and confusing it, then filters - map - resynthesize (ie, use the plugin rather than the script), and have the "tilable" options disabled since they tend to grab samples from the opposite edge of the image (if I want a tilable image, I'll use the tiling filter separately...)

Probably the biggest factor for simple success is to have the object you want to remove be on its own (surrounded on all sides by similar textures) -- if it isn't, then you need to do things the long way -- eg, if you want to remove the leftmost wheelchair from this image, and you want it to be replaced by grass when three of its borders are touching non-grass, then you'll find that it ends up somewhat messy since it attempts to merge four different edge textures. In this case you'll need to copy a section of your desired fill texture (ie, a rectangle of pure grass) into a separate image (specifically, a single layer image with no transparency); then on the original image select the object to remove, open resynthesizer, and select the "fill texture" image as the texture source; this way the generated texture will both match the surroundings of the original as much as possible, while being filled with the "surroundings" that you've specifically chosen. Having taken a sample of "pure grass" and a sample of "pure stone", then removing the top and bottom halves of the wheelchair with each respectively, the results are nicer. (with the exception that the first two images were produced with a mouse and twenty minutes of careful selecting, and the final one was 5 minutes work with a laptop nipple, so there are still some bits of wheelchair poking out of the sides...)

Incidentally, does photoshop have SIOX yet? Having the features "vaguely scribble in the general area of an object to have the object selected precisely" and "automatically and realistically remove a selection" could potentially combine to form "one-click realistic object removal" \o/

Just tried expanding a panorama as in the demo too, and resynthesizer does that pretty well too (please ignore the fact that the source material is REALLY shittily exposed, it was shot on a mobile phone with no manual exposure mode :( ):

original
expanded

Just tried expanding a panorama as in the demo too, and resynthesizer does that pretty well too (please ignore the fact that the source material is REALLY shittily exposed, it was shot on a mobile phone with no manual exposure mode :( ):

original
expanded

The demo seems to have been done under pretty much ideal conditions -- when I've been working with similarly suitable source material, I've had similarly awesome results from resynthesizer.

Some examples of things I've resynthesized in the last 10 minutes for the sake of demonstration, mixed in with some "missing object" quiz questions I made for an anime society a couple of months ago, again done with the resynthesizer: http://shishnet.org/ufufuf/mo/

This is what I get using the plugin on its own: http://shishnet.org/ufufuf/resynth2.jpg

Do note that the script-fu wrapper works better for larger images, which this isn't

Also, the example from the video, done with gimp instead, the results are pretty similar (IMO, better, but I'm pretty sure that the "improvements" are just luck): http://shishnet.org/ufufuf/panorama-synth.png

Having been using the resynthesizer for years, I've developed a knack for which source images will work well and which won't, and the thing that struck me about that video was that the source images are pretty much ideal conditions -- I'll be impressed when they can get good results on the images that aren't so clean :-)

This is what I get using the plugin on its own: http://shishnet.org/ufufuf/resynth2.jpg

Do note that the script-fu wrapper works better for larger images, which this isn't

Also, the example from the video, done with gimp instead, the results are pretty similar (IMO, better, but I'm pretty sure that the "improvements" are just luck): http://shishnet.org/ufufuf/panorama-synth.png

Having been using the resynthesizer for years, I've developed a knack for which source images will work well and which won't, and the thing that struck me about that video was that the source images are pretty much ideal conditions -- I'll be impressed when they can get good results on the images that aren't so clean :-)

Too bad the EU saw fit to punish all of its citizens by making them go dig up a browser somewhere.

I got bored of explaining why this is wrong over and over, so here's a link

Slightly tangential, but having sent several bug reports regarding ugly interfaces, I ponder tying them together in something like a constructive criticism blog; some random recent examples being an official KDE4 screenshot showing how "good" it looks, the default mythbuntu theme (yes, all these styles are a single theme), and some things which are just plain ugly.

Good idea / bad idea? Would anybody benefit from a collection of "what not to do"'s? Anybody interested in writing up some criticisms of their own?

Slightly tangential, but having sent several bug reports regarding ugly interfaces, I ponder tying them together in something like a constructive criticism blog; some random recent examples being an official KDE4 screenshot showing how "good" it looks, the default mythbuntu theme (yes, all these styles are a single theme), and some things which are just plain ugly.

Good idea / bad idea? Would anybody benefit from a collection of "what not to do"'s? Anybody interested in writing up some criticisms of their own?

Slightly tangential, but having sent several bug reports regarding ugly interfaces, I ponder tying them together in something like a constructive criticism blog; some random recent examples being an official KDE4 screenshot showing how "good" it looks, the default mythbuntu theme (yes, all these styles are a single theme), and some things which are just plain ugly.

Good idea / bad idea? Would anybody benefit from a collection of "what not to do"'s? Anybody interested in writing up some criticisms of their own?

You come up with a neat idea, then discover that someone more famous than you had a similar idea, and spend the rest of your life being called a copy cat :(

I can't help but wonder if beastie's appearance means that BSD is powering condom machines...

A couple of days ago some guy got flamed for saying "The alignment is off, doesn't anybody even look at their software before releasing it?", with the most useful response being "your font settings are probably different to the developer's, they don't see what you see"; and I agreed with them. But looking at screenshots for myself, even the official screenshots showing how good it looks, look bad. annotated example. (PS. Any idea where I can send that to to have people fix it?)

/me goes back to enlightenment 17, ever more appreciative of Raster's perfectionism...

That looks like an angry OAP; For real bad-assery, you really need a truck

Fast forward = right arrow, pause = space bar; the rest of the key bindings I frequently use are shown here -- to clarify, my point was that sensible key bindings > non-standard graphical widgets

I left out a load of cool things because I knew that BeOS had them first, didn't know about that one though...

It's incredibly useful when you run a shared server with many not-completely-trusted users

More windows allows you to see more, but it gets messy -- currently I have one xterm for communications (irssi, mutt, slrn); one media (mpc, mplayer); one for each project I'm working on 4 * (vim, database prompt, root prompt, shell); which would be 19 unorganised xterms rather than 6, and with 6 I can keep the 3 I'm interested in on top of the screen at all times and never need to switch.

This isn't even getting to the other features of screen -- such as being able to detach and reattach elsewhere, or from two places at once -- this allows me to run my communications screen from a server which is online 24/7, and I can connect to it from home, work, uni, etc and have everything just how I left it. Multiple attachments also allows several people to see what someone's doing, which is useful when doing dangerous work on critical systems -- you can have someone look over your shoulder and yell "don't do that!", even when that other person is in another country.

Ack, I'd tried to repress that from my memory... Yeah, mplayer has a terrible GUI, but it's optional -- which is why I use the command line on every OS. You can either type, or drag & drop a video file onto the .exe, and it just plays, with no fuss. Further interaction is done purely by keyboard, and the keybindings are flexible enough to do pretty much anything~ And before going "ew, command line!", note that I've recommended the "drag & drop onto the .exe" approach to several generic windows users, and they've all thanked me for introducing them to such an awesome media player :P

Filesystem in User Space; it makes developing new filesystems ridiculously easy, leading to things like wikifs -- you can mount wikipedia as any other disk drive, then the wiki articles appear as text files which you can edit with any standard editor.

Just a few weeks back there was some anime / subliminal propoganda sponsored by the japanese equivalent of NASA, and they had suits which looked just like that :O

(That series also introduced me to reverse polish calculators, and it's true, I can no longer stand to use a regular calculator; RPN just seems so much more elegant...)

Just a few weeks back there was some anime / subliminal propoganda sponsored by the japanese equivalent of NASA, and they had suits which looked just like that :O

(That series also introduced me to reverse polish calculators, and it's true, I can no longer stand to use a regular calculator; RPN just seems so much more elegant...)

Just a few weeks back there was some anime / subliminal propoganda sponsored by the japanese equivalent of NASA, and they had suits which looked just like that :O

(That series also introduced me to reverse polish calculators, and it's true, I can no longer stand to use a regular calculator; RPN just seems so much more elegant...)

You know you've been programming too long when you turn your code comments into ASCII space tanks (about half way down). It won't get through the lameness filter though :(

On topic: if that comment only needed 79 characters, then 80 should be plenty! :P

I had much the same problem, which I attempted to solve with danbooru; after 4 hours of banging my head against ruby on rails and postgres, I decided that there needed to be something which was much like danbooru but a lot easier to set up. Being unable to find such a thing, I made one for myself, and the result is here. (tech info / downloads / etc)

I had much the same problem, which I attempted to solve with danbooru; after 4 hours of banging my head against ruby on rails and postgres, I decided that there needed to be something which was much like danbooru but a lot easier to set up. Being unable to find such a thing, I made one for myself, and the result is here. (tech info / downloads / etc)

On that topic, my own hack of choice; an animated presentation engine in 341 bytes, a protest against my classmates' multi-megabyte powerpoint based presentations:

$c=`clear`;foreach(`tail -c+341 $0|zcat`){if(/^p/){print@l;@l=()} elsif(/^d(.+)/){select$u,$u,$u,$1/10}elsif(/^s(.)/ ){$s=10;$e=-30; $m=-1;if('o'eq$1){$s=-30;$e=10;$m=1}for($i=$s;$i!= $e;$i+=$m){$j=0; foreach(@l){print substr($_,(($i+$j++)>0?($i+$j)**2:0),-1)."\n"} select($u,$u,$u,0.1);print$c}}elsif(/^c/){print$c; @l=()} else{push@l,$_}}__END__

The full presentation with data file; a ~1:30 demonstration on floating point binary in under 2KB.

Source with generator; have p7zip installed and run "./build.sh floating" to generate the presentation.

I don't do much perl hackery, I'm sure someone who does could do shorter ;)