- Cin - Cinelerra-GG Mailing Lists

Test
by Sam 27 Jul '19

27 Jul '19

Just a test --

1 0

FFmpeg 4.2 branched
by Andrew Randrianasulu 24 Jul '19

24 Jul '19

Hi, all! https://git.ffmpeg.org/gitweb/ffmpeg.git/shortlog/refs/heads/release/4.2 http://ffmpeg.org/pipermail/ffmpeg-devel/2019-July/246897.html

1 0

Buttons for Mask tool window
by Igor BEGHETTO 19 Jul '19

19 Jul '19

Mainly for GG (... but also for the others) I added a file compressed and an animated GIF (photomontage with the buttons) to show the layout or an idea of layout. In the compressed file You find: 1) the "6" buttons for Mask Linear and Smooth, for each theme. They are always the first version I have done with the blue line in smooth buttons (due at my time). Sorry Olaf, I did not find your SVG files in Cakewalk theme to derivate the new buttons from your original buttons, but the buttons I have done have the right colour. For the SUV theme You find the "3" buttons for the Scale features. Why are there six PNGs instead of five? And, as first version and testing, You find the Presets (7 btns) and Positions (2 btns) buttons for only SUV theme. I hope that these buttons work like Mask Linear/Smooth buttons: every button use 4 icons, with three common icons ("mask_button_dn.png", "mask_button_hi.png", "mask_button_up.png"). If You (GG) think different, let me know, please. Sorry Sam, I have changed again a few of them. After some test with SUV theme we (all) can decide, I think. @GG and @Olaf More, I noticed two things about Neophyte and Cakewalk themes: A) Neophyte theme >> neophyte.C - code line 361 ("reset_dn.png") inverted with line 362 ("reset_hi.png") It should be: /* ****************** */ new_button("reset.png", "reset_up.png", "reset_hi.png", "reset_dn.png", "reset_button"); /* ****************** */ The PNGs files in theme_neophyte/data/PNGs are right but they are an old version. You can find the new PNGs files for reset button, in the compressed file, with the others mask buttons. (If/When Olaf wants he can update with one of his version) B) Cakewalk theme >> cakewalk.C - code line 272 ("reset_dn.png") inverted with line 273 ("reset_hi.png") I think, the codes should be (for consistency with the others codes/themes) like in point A, and in theme_cakewalk/data/ the two PNGs files have to be renamed: you have to swap the two names "reset_dn.png" with "reset_hi.png". Thank you so much! I go to drink something. (too many buttons) ;-) IgorBeg (Cin version: "cinelerra-5.1-ub16-x86_64-static_20190711-testing")

3 8

Website & MantisBT
by Sam 14 Jul '19

14 Jul '19

Hi guys, The work on the server is finished. MantisBT was affected by some work and therefore you may not have had the full functionality. More maintenance work will be done next weekend to further improve our website. Sam --

1 0

Re: [Cin] [CinCV TNG] about Cin-GG
by matheu 14 Jul '19

14 Jul '19

Hi! Sam: > Hi Matheu, > > I'm sorry you had trouble subscribing to the mailing list. > > I can disable all the security mechanisms, but as a result we have more > work with all the manual cleanups caused by spammers. We are exposed to > many spam attacks every day and use existing protection mechanisms to > protect ourselves and our community from such attacks. Here are some > numbers about spam, last weeks 2786 attacks were blocked based on IP > address. 17 direct login attempts with blacklisted email addresses have > been blocked. 7 brute force attacks and 70 complex attacks. All > attackers want to remain anonymous. However, blacklisting Tor exit nodes is usually not doing anything to the attackers, because of the way Tor works. E.g., one of the addresses that you blacklist is: http://metrics.torproject.org/exonerator.html?ip=185.220.101.27&timestamp=2… a Tor exit node with probably heavy traffic of all kinds of people, most of them probably legitimate visitors. The malicious ones, esp. if they are expert or just more advanced, your ban does pretty much nothing to them, because they can do so much more than just change identity in the top right corner of Tor browser (which means change who you appear to be to the website that you are visiting, i.e. the exit node)... So banning an exit node, which is what the CleanTalk probably does (no time to investigate, but I think I saw some address with that name in some instances when I tried visiting https://www.cinelerra-gg.org and was banned, is just wrong thing to do... Also there are FOSS captcha's available, way more benign to users, but I have no time to search for links to those... > I also prefer to surf the Internet > anonymously as far as possible, but we use these protection techniques > to protect ourselves and not to use our visitors for advertising > purposes. These security tools help us spend our time on more important > things, such as more time improving Cinelerra. I would love to do > without such tools, but unfortunately this would make dynamic content of > our users impossible. > > I can add you manually. I'll send you an email But whatever you use to counter spam and attacks is blacklisting exit nodes that I used today as well. [...] > Sam [...]

2 5

Fixed setting: MantisBT 403 error
by Sam 12 Jul '19

12 Jul '19

FYI: When viewing images in the Mantis Bugtracker, you sometimes got the error message 403 "Access denied". This setting error has been fixed. Sam --

2 2

Checkin includes ffmpeg upgrade to 4.1.4
by Phyllis Smith 11 Jul '19

11 Jul '19

Just an FYI that today's GIT checkin includes an upgrade of ffmpeg to 4.1.4 (previously was 4.1). More Masking changes also checked in -- refer to MantisBT issue #2. gg/Phyllis

1 0

Mask tool: everything to be done again!
by Andrea paz 10 Jul '19

10 Jul '19

[I'm sorry, I'm kidding :-) ] The mask tool has become really useful and powerful. I thank Sam and IgorB for the many inputs and tests; Phyllis for keeping GG calm [ ;-) ] and everything else; and GG because he is an incredible coder. CinGG is a great project because you are in charge of it (not to mention Pierre, MatN, Andrew-R, einhander, Olaf and many others...).

2 1

Questions about compiling
by Andrea paz 05 Jul '19

05 Jul '19

- To compile new releases from source, I use the following ./configure: ./configure --prefix=/usr --with-jobs=8 --without-oss --with-single-user –with-booby But I get warnings I don't understand: [paz@archpaz cinelerra-5.1]$ ./configure --prefix=/usr --with-jobs=8 --without-oss --with-single-user –with-booby ... configure: WARNING: you should use --build, --host, --target configure: WARNING: invalid host type: –with-booby ... What are the best ./configure options to use for compiling? - Another question: I noticed the following warning when compiling. It is a warning that repeats itself dozens of times before continuing: ... ./libavutil/mem.h:342:1: warning: ‘alloc_size’ attribute ignored on a function returning ‘int’ [-Wattributes] 342 | av_alloc_size(2, 3) int av_reallocp_array(void *ptr, size_t nmemb, size_t size); | ^~~~~~~~~~~~~ ... Everything's working fine, so I don't think it's important. PS: Great idea the smooth button in the mask tool! PPS: Have you heard of the new -O1g instead of -O1?

2 2

Typo in recent commit?
by Andrew Randrianasulu 03 Jul '19

03 Jul '19

In https://git.cinelerra-gg.org/git/?p=goodguy/cinelerra.git;a=blobdiff;f=cine… index c00ca73..dbeee00 100644 (file) --- a/cinelerra-5.1/cinelerra/performanceprefs.C +++ b/cinelerra-5.1/cinelerra/performanceprefs.C @@ -332,8 +332,8 @@ PrefsUseHWDev::PrefsUseHWDev(PreferencesWindow *pwindow, void PrefsUseHWDev::create_objects() { BC_PopupTextBox::create_objects(); - set_tooltip(_("vdpau - Nvidia, Nouveau, Amdgpu\n" - "vaapi - Broadcom, Intel HD graphics, Radeon\n" + set_tooltip(_("vdpau - Nvidia, Nouveau\n" + "vaapi - Broadcom, Intel HD graphics, Amdgup, Radeon\n" "cuda - Nvidia + Cuda SDK")); hw_dev_names.append(new BC_ListBoxItem(_("none"))); #ifdef HAVE_VAAPI I think It still should be "amdgpu", not "amdgup" (this is just tooltip, yet ..) I think vdpau decode also availiable on radeon/radeonsi, but this fact depend on state trackers enabled during mesa compilation: LANG=C ls -la /usr/X11R7/lib/*vdpau total 34964 drwxr-xr-x 2 root root 4096 Jul 1 07:00 ./ drwxr-xr-x 15 root root 20480 Jul 1 09:58 ../ -rwxr-xr-x 1 root root 1352 May 17 2014 libvdpau_nouveau.la* lrwxrwxrwx 1 root root 25 Jul 1 07:00 libvdpau_nouveau.so -> libvdpau_nouveau.so.1.0.0* lrwxrwxrwx 1 root root 25 Jul 1 07:00 libvdpau_nouveau.so.1 -> libvdpau_nouveau.so.1.0.0* lrwxrwxrwx 1 root root 25 Jul 1 07:00 libvdpau_nouveau.so.1.0 -> libvdpau_nouveau.so.1.0.0* -rwxr-xr-x 4 root root 8919356 Jul 1 07:00 libvdpau_nouveau.so.1.0.0* lrwxrwxrwx 1 root root 22 Jul 1 07:00 libvdpau_r300.so -> libvdpau_r300.so.1.0.0* lrwxrwxrwx 1 root root 22 Jul 1 07:00 libvdpau_r300.so.1 -> libvdpau_r300.so.1.0.0* lrwxrwxrwx 1 root root 22 Jul 1 07:00 libvdpau_r300.so.1.0 -> libvdpau_r300.so.1.0.0* -rwxr-xr-x 4 root root 8919356 Jul 1 07:00 libvdpau_r300.so.1.0.0* -rwxr-xr-x 1 root root 1339 May 17 2014 libvdpau_r600.la* lrwxrwxrwx 1 root root 22 Jul 1 07:00 libvdpau_r600.so -> libvdpau_r600.so.1.0.0* lrwxrwxrwx 1 root root 22 Jul 1 07:00 libvdpau_r600.so.1 -> libvdpau_r600.so.1.0.0* lrwxrwxrwx 1 root root 22 Jul 1 07:00 libvdpau_r600.so.1.0 -> libvdpau_r600.so.1.0.0* -rwxr-xr-x 4 root root 8919356 Jul 1 07:00 libvdpau_r600.so.1.0.0* -rwxr-xr-x 1 root root 1363 May 17 2014 libvdpau_radeonsi.la* lrwxrwxrwx 1 root root 26 Jul 1 07:00 libvdpau_radeonsi.so -> libvdpau_radeonsi.so.1.0.0* lrwxrwxrwx 1 root root 26 Jul 1 07:00 libvdpau_radeonsi.so.1 -> libvdpau_radeonsi.so.1.0.0* lrwxrwxrwx 1 root root 26 Jul 1 07:00 libvdpau_radeonsi.so.1.0 -> libvdpau_radeonsi.so.1.0.0* -rwxr-xr-x 4 root root 8919356 Jul 1 07:00 libvdpau_radeonsi.so.1.0.0* may be splitting hw accel into decode/encode/filtering still makes some sense, as mid-term plan? (so, you decode on Intel, and encode on NVidia, for example, if you have two cards)

2 1