Re: More LiveCD space optimizations
Hello Martin, Martin Owens [2010-11-07 1:36 -0400]: > That already is written in C, it's the script that pulls in the config > and runs the usb_modeswitch program which is written in tcl. > > It should be very possible to convert it to python or vala. No python please, it has the same poor boot time behaviour. C or vala or anything compiled should do. > I have to wonder what 200 udev rules all with different vendor and > product ids does to the boot time. Not much, since either none or just one will match on your system. usb-modeswitch needs those long lists, but udev is good at efficient rule matching and parsing (that's what it is for, after all). Martin -- Martin Pitthttp://www.piware.de Ubuntu Developer http://www.ubuntu.com Debian Developer http://www.debian.org -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On Fri, 2010-11-05 at 16:20 -0400, Martin Pitt wrote: > One thing that currently needs it is usb-modeswitch. I'd love the > usb-modeswitch-dispatcher thing to be rewritten in C, Vala, or another > compiled language. Not only is it holding tcl in the default install, > but it also dramatically slows down boot. That already is written in C, it's the script that pulls in the config and runs the usb_modeswitch program which is written in tcl. It should be very possible to convert it to python or vala. I have to wonder what 200 udev rules all with different vendor and product ids does to the boot time. Martin, -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On 10/08/2010 04:54 AM, Matthias Klose wrote: > [ compression related discussion removed ] > > So maybe we can save some MB with better compression, but we can save more by > not including files at all. Of course this requires inspection of the > packages > included on the liveCD. In the past we did identify some issues and did add > some diagnostics to the live CD build logs [1]. Of course you can't run > anything and lengthen the live CD build, but some additional diagnostics > maybe > could be run. > > In the past we did see wasted space: > > - firefox and xulrunner shipping duplicate .js files > Well, Firefox is no longer built on top of xulrunner, so this is necessary, especially with the PGO optimizations if we can get them. If webkit has sufficient accessibility and we can port yelp to webkit, we can drop xulrunner from the CD. -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
John McCabe-Dansted [2010-10-08 0:07 +0800]: > We could test each file to ensure the image is identical, perhaps > using pngtopnm, and md5sum. This would be especially important for > jpegrescan/jpgcrush, which is at version 0.0.0-1. I use a simple test script for this kind of check, see https://bugs.launchpad.net/ubuntu/+source/advancecomp/+bug/671599/comments/1 Martin -- Martin Pitt| http://www.piware.de Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org) -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
Matthias, Matthias Klose [2010-10-08 11:54 +0200]: > - Packages which should not be on the CD. Some things should not be > on the CD at all. Looking at the current live CD log, a typical > candidate for this would be tcl8.4. Why is it there, and how can > it be avoided? One thing that currently needs it is usb-modeswitch. I'd love the usb-modeswitch-dispatcher thing to be rewritten in C, Vala, or another compiled language. Not only is it holding tcl in the default install, but it also dramatically slows down boot. > - Localized help images. You cannot just remove the images from an > application's help, but in the past we did ship all these localized > help images on the CD. CC'ing Martin, don't know the current status. > However it looks like there are some xml files which maybe should > be part of the language packs. Since Lucid (or so) we strip those out of the app packages and ship them in the language packs. That already saved us a lot of space. Martin -- Martin Pitt| http://www.piware.de Ubuntu Developer (www.ubuntu.com) | Debian Developer (www.debian.org) -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On Fri, Oct 8, 2010 at 2:01 AM, Matthias Klose wrote: >> Also, there are 12MB of jar files, which are basically zip files. We >> can also shrink those by 5MB or so with advzip, but that doesn't seem >> to shrink a .tgz of them so it may not shrink the liveCD. Since zip >> files compress file by file, we may be able to save space on the >> liveCD by running "advzip -z -0" on them. That would expand them to >> 24MB, but reduces the size of a .tgz of them to 4.6MB, possibly saving >> space on the liveCD if squashfs is similarly efficient. > > how does OOo behave with the repacked zip file? is it faster, slower, does it No, I made a script to open oowriter 100 times. It didn't find any consistent difference in performance. > need more memory when it runs? imo, changes like this should be integrated > into gzip needs less than 1MB to decode (or even encode). The effect on memory usage is likely to be minimal. http://tukaani.org/lzma/benchmarks.html > the package build process, and sent upstream. patches welcome. > > same for jar files. are these extracted as fast as without your changes by the > jvm? if not, then these should be left alone (and afaik there shouldn't be any > jar files on the live CD). FYI, most of the jar files come from firefox and openoffice. Firefox refuses to start without these jar files. I doubt they are used by a jvm. Using the 7z deflate instead of gzip shouldn't harm decompression time. In fact, it should improve speed slightly because there is less compressed input to parse (If I tar up /etc and compress it with gzip and advdef, the one compressed with advdev does in fact seem to gunzip very slightly faster). The other question is does compressing decompressed jars (or visa-versa) affect performance. A atom based netbook with a rotational disk seemed like a good machine to test on as it is towards the low end of performance. Repeatedly running oowriter and firefox ten times did not lead to and consistent performance differences (see attaced timeopen.ar). If there is any difference it would be within a few percent. This suggests that we can decompress them (to save liveCD space), or compress them (to save installed space) without having much effect on performance. It is also plausible that decompressing the jars saves installed space on 'Btrfs -o compress' filesystems, but I have not tested whether Btrfs compression heuristics automatically detect the jars as being compressible. -- John C. McCabe-Dansted timeopen.ar Description: Binary data -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
Sorry for the 4th post in a row, but I added a script that uses AdvanceCOMP to recompress the .gz files that aren't man pages, and I had to share my findings. AdvanceCOMP? | ISO size (B) | Install (KiB) No | 711,032,832 | 2,474,660 Yes | 707,821,568 | 2,469,568 --- Savings |3,211,264 | 5,092 The script is attached. Due to ext4 extent allocation and the order of the files on the CD, the reordered CD made by 98make-disc boots faster, but its installed size is 180 MB bigger, so this new mksquashfs ordering (in 98make-disc) is a tradeoff. This new ordering is not used in the actual CD building process, though I filed a bug for it [1]. Should I revert to the default ordering done by mksquashfs or start using ext3 installations to compensate, for testing? - Louis [1] https://bugs.launchpad.net/ubuntu/+source/livecd-rootfs/+bug/589629 92gz-optimisation-experimental Description: Binary data -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
Apologies for the previous attachment, it didn't have the addition for man-page symbolic links. I attach the proper one this time. - Louis ubuntu-opt.tar.gz Description: GNU Zip compressed data -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
2010-10-07 16:07 GMT John McCabe-Dansted : > On Thu, Oct 7, 2010 at 10:05 AM, Louis Simard wrote: >> Do you want me to add to my script any of the optimisations discussed >> in your email? They are: Using AdvanceCOMP to recompress .png images >> and gzipped files; using either of jpegoptim or jpegrescan to >> losslessly recompress .jpg images; "transcoding" man pages from .gz to >> .lzma. I'm not going to add untested optimisations yet, such as >> transcoding *all* .gz files to .lzma. > > Sure. This could help with testing that these actually work ;). It works! :) 'man' reads its files correctly, after an addition to the script to fix the broken symlinks (for example, zfgrep.1.gz pointing to zgrep.1.gz in /usr/share/man/man1 which became zgrep.1.lzma), OpenOffice.org opens the Human icons correctly, all of the PNG images compare equal using pngtopnm (although some emit a warning about the pixel aspect ratio, i.e. non-square pixels) and Java is able to read .jar files that have been recompressed. However, while recompressing the files helped the ISO size, it made the install size grow by about 180 MB... When | ISO size (B) | Install (KiB) Old | 718,864,384 | 2,293,740 New | 711,032,832 | 2,474,660 Ow! What's going on? Here's the methodology I used for that result. * The base CD is Ubuntu 10.04.1. * Using the virtual hardware provided by VirtualBox OSE 3.1.6 r59338, 384 MB RAM, 5 GB SCSI hard drive and IDE CD-ROM drive using the ISO. * Look at the size of the CD ISO using ls -l. * Using 'df', look at the number of disk blocks used on the VM's hard drive before rebooting from an installation via the GUI: a) Language: English b) Time zone: Europe/United Kingdom Time c) Keyboard layout: USA d) Disk space preparation: Manual partitioning, no swap, fill the entire drive with an ext4 partition for / e) Username, password and computer name: "dummy" Attached are the scripts I used to mount the original CD, recompress the files and make the new CD. No CDs were harmed in the making of this email. - Louis ubuntu-opt.tar.gz Description: GNU Zip compressed data -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
2010-10-08 09:54 GMT Matthias Klose : > In the past we did see wasted space: > > - Packages which should not be on the CD. Some things should not be > on the CD at all. Looking at the current live CD log, a typical > candidate for this would be tcl8.4. Why is it there, and how can > it be avoided? foo2zjs made APT install that package. $ aptitude why tcl8.4 i tk8.4 Depends tcl8.4 (>= 8.4.16) $ aptitude why tk8.4 i foo2zjs Recommends tk8.4 I'm sure there will be other examples, though I'm not as familiar with the LiveCD's packages as you guys at Canonical. -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
[ compression related discussion removed ] So maybe we can save some MB with better compression, but we can save more by not including files at all. Of course this requires inspection of the packages included on the liveCD. In the past we did identify some issues and did add some diagnostics to the live CD build logs [1]. Of course you can't run anything and lengthen the live CD build, but some additional diagnostics maybe could be run. In the past we did see wasted space: - Packages which should not be on the CD. Some things should not be on the CD at all. Looking at the current live CD log, a typical candidate for this would be tcl8.4. Why is it there, and how can it be avoided? - Large doc directories. If a package becomes too large, maybe it is worth to split a package into foo and foo-doc, and not ship foo-doc on the CD (yes there are other ideas not to ship doc dirs at all). See python-couchdb for an example. The API documentation does not need to be on the live CD. The same may be true for other python packages. - Localized help images. You cannot just remove the images from an application's help, but in the past we did ship all these localized help images on the CD. CC'ing Martin, don't know the current status. However it looks like there are some xml files which maybe should be part of the language packs. - Duplicate files. While this is not that important on the live CD, it's important for the alternate CDs. Looking at the list of duplicate files, I see a lot of potential in: - all the mono packages and libraries - broken build systems shipping doc files in every binary package. see the upstream changelog.gz files (e.g. gnome and OOo). - firefox and xulrunner shipping duplicate .js files - package specific stuff (libc6-dev having some identical libs in /usr/lib/xen). - you may see and find more if you are familiar with a particular package. There is potential in saving space with better compression, but IMO you can even save more with closely looking what goes on the CD (where we currently don't do a good job). The good thing is that both approaches don't exclude each other. Matthias [1] http://people.canonical.com/~ubuntu-archive/livefs-build-logs/maverick/ubuntu/latest/livecd-20101007-i386.out -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On 10/08/2010 12:22 AM, Louis Simard wrote: >> There are a over a dozen different types of file to be tested (and >> there may be more than one application that wants to read them). For >> reference, I have attached them. Probably the most important thing to >> check is that printing still works, as many of the gz files seem to >> e.g. ppd files. >> >> Maybe if you added it to your script and just gave the resulting iso a >> spin in a VM to see if there was obvious breakage? > > I have no printer supported by OpenPrinting PPDs to test this with, > but a VM is exactly what I used to test SVG, XML and PNG optimisations > in May (and realise that librsvg had a bug that needed worked around > in Scour! [librsvgbug]). I'll do this, but PPDs would still need > testing afterwards. > > A separate thread and perhaps contact people already exist for the PPD > gzip compression ([openprinting-ppds-gzip]), and perhaps it would be > best to communicate with these people to have them test and add > AdvanceCOMP to their gzipping. > The space occupation of the PPDs I have already solved in Maverick. Most of the PPDs in /usr/share/ppd and also the Foomatic XML data in /usr/share/foomatic are replaced by highly compressed PPD file archives based on LZMA (in /usr/lib/cups/driver). This saves around 30-40 MB on the live system even getting the PPDs of the former openprinting-ppds-extra package onto the live CD. The tools for building these archives are from a Google Summer of Code project which I have mentored for OpenPrinting. See https://bugs.launchpad.net/bugs/493282 and http://pypi.python.org/pypi/pyppd The packages on the CD which have their PPDs compressed now are: foomatic-db, openprinting-ppds, hplip-data, splix Till -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
* LONG MESSAGE WARNING * While I've tried to reduce the quotes and quote nesting as much as I could, this message is still long. It is still important to read, when you have time. 2010-10-07 16:07 GMT John McCabe-Dansted : > On Thu, Oct 7, 2010 at 10:05 AM, Louis Simard wrote: >> > > I think this will be discussed at UDS-N, see: > http://archives.free.net.ph/message/20101004.065026.e553efd1.en.html Awesome! Will a digest of this conversation need to be posted to ubuntu-devel only once done, continuing on ubuntu-devel-discuss for now? >> 2010-10-06 16:08 GMT John McCabe-Dansted : >>> [...] I note that we can save further space by: >>> >>> 1) Using advdef on the png files in addition to optipng. This is what >>> optimizegraphics does, and this shrinks the pngs on the Maverick RC >>> liveCD from about 100.1MB to 85.3MB providing a saving of 14.8MB. > > We could test each file [after using advpng on them] > to ensure the image is identical, perhaps > using pngtopnm, and md5sum. This would be especially important for > jpegrescan/jpgcrush, which is at version 0.0.0-1. Good idea. I may be able to integrate this test into my script as an option. >>> 2) Recompressing gz files with advdef. Using advdef, we can shrink the >>> gz files from 89.5MB to 84.8MB, [...] a saving of 4.7MB. >> >> [...] I did use 7zip's Deflate compressor to recompress a >> .zip file of OpenOffice.org's from 5.9 MB to 5.4 MB. [...] > > You mean images_human.zip? Yes, thanks. :) I had forgotten the name. > I have a hunch that compressing that file > wouldn't actually save space on the liveCD as I can gzip it down to > 3.9MB. It may be better to leave it as an uncompressed zip, and let > squashfs deal with it. Per that "Performance - Disk footprint" thread from ubuntu-devel [brainstorm], we may actually want to also care about the installed size, and use the 7zip recompression. While it's not going to be *perfectly optimal*, reducing both the CD footprint and the installed size by 0.5 MB using 7zip sounds better than reducing the CD footprint by 2 MB, but increasing the installed size by more than 2 MB. And if you managed to re-gzip the zip, squashfs will also manage to re-lzma the zip for more savings and still a decent installed size. You should test this again with lzma, I think. > Recompressing the pngs contained in the zip > sounds worthwhile though. Strangely, even running advzip -z -0 > images_human.zip shrinks it by 3%, and even shrinks the corresponding > images_human.zip.gz file I believe you there, only because the original situation has a deflated container (png) within another deflated container (zip). Counter-intuitive, but something to consider. > Also, there are 12MB of jar files, which are basically zip files. We > can also shrink those by 5MB or so with advzip, but that doesn't seem > to shrink a .tgz of them so it may not shrink the liveCD. Since zip > files compress file by file, we may be able to save space on the > liveCD by running "advzip -z -0" on them. That would expand them to > 24MB, but reduces the size of a .tgz of them to 4.6MB, possibly saving > space on the liveCD if squashfs is similarly efficient. > same for jar files. are these extracted as fast as without your changes by the > jvm? if not, then these should be left alone (and afaik there shouldn't be any > jar files on the live CD). Aha! I completely forgot .jar files. The OpenJDK package itself may become much smaller after this, because of the huge runtime rt.jar. Must test and benchmark this! I believe OpenOffice.org is a huge user of Java, so there would be .jar files on the LiveCD from that too. >>> A further 10MB could be saved by recompressing the gz files as lzma. >> At what LZMA compression level? Default (7) or --best (9)? > --best I just want to add that blanket recompression of gzip files as lzma with --best could be harmful, but with small files it's probably OK. LZMA uses a huge dictionary to do its work, which needs to be allocated even on the decompressing side, and --best may overrun the memory of low-end computers on larger files. > Also, if we want to take replacing deflate with lzma to extremes, we > could replace the deflate compression in the png files with lzma. A > command that does this is "advpng -z -0 $f && lzma --best $f". I found > that this could save 18.7MB. However, It may also degrade performance > slightly, but I doubt it would be too significant on modern CPUs. > Running unlzma on all 66MB of the .png.lzma files takes: > real 1m2.666s > user 0m6.540s > sys 0m5.610s > > I think the user/sys are the relevant ones, and taking 12s to read > every png doesn't seem too bad. The main thing is that I doubt that it > would work out of the box. > > If we use lzma in the squashfs, just deflating them all with advpng -z > -0 could reduce the liveCD size. Probably wouldn't help the installed > size though. Indeed. > There are a over a dozen different types of file to be tested (and > there may be more th
Re: More LiveCD space optimizations
2010-10-07 16:29 GMT Martin Owens : > On Fri, 2010-10-08 at 00:07 +0800, John McCabe-Dansted wrote: >> Strangely, even running advzip -z -0 >> images_human.zip shrinks it by 3%, and even shrinks the corresponding >> images_human.zip.gz file > > That's not strange, that's just entropic packing principles. You've got > a bunch of assumptions that can be made about data and a bunch of > compression iterations, each make assumptions about the nature of the > data and some are fitting together better. > > I'm keen on this work since saving space allows for all sorts of > goodies. Did we save space with any of the SVG cleaning or did that need > to be brought up to the packaging level? > > Martin, > > Back in May, the preliminary testing I did on the LiveCD's .svg files resulted in the finding that using Scour on them saved about 7 MB [1]. Of course, not only the LiveCD's packages use .svg files, and it would be important to get that to other packages as well, for download times/bandwidth use, if for any other reason. Perhaps rendering speed would increase too, in SVG's case, but the other file formats discussed in this thread have different characteristics. So it needed to be brought up at the packaging level [1]. Scour will probably itself need to be packaged too, to be included as build-depends for packages that have SVG files (which is a lot of application packages, since most have an SVG icon) to work well with 'apt-get source'. [1] https://lists.ubuntu.com/archives/ubuntu-devel-discuss/2010-May/011505.html -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On 07.10.2010 18:07, John McCabe-Dansted wrote: >> That's an interesting optimisation; I didn't really know about it >> either. However, I did use 7zip's Deflate compressor to recompress a >> .zip file of OpenOffice.org's from 5.9 MB to 5.4 MB. The method was >> rather crude, but it did the job: >> >> mkdir extracted >> cd extracted >> unzip ../file.zip >> 7z a -tzip -mx=9 -mfb=258 file.repack.zip extracted/* >> rm -r extracted > > You mean images_human.zip? I have a hunch that compressing that file > wouldn't actually save space on the liveCD as I can gzip it down to > 3.9MB. It may be better to leave it as an uncompressed zip, and let > squashfs deal with it. Recompressing the pngs contained in the zip > sounds worthwhile though. Strangely, even running advzip -z -0 > images_human.zip shrinks it by 3%, and even shrinks the corresponding > images_human.zip.gz file > > Also, there are 12MB of jar files, which are basically zip files. We > can also shrink those by 5MB or so with advzip, but that doesn't seem > to shrink a .tgz of them so it may not shrink the liveCD. Since zip > files compress file by file, we may be able to save space on the > liveCD by running "advzip -z -0" on them. That would expand them to > 24MB, but reduces the size of a .tgz of them to 4.6MB, possibly saving > space on the liveCD if squashfs is similarly efficient. how does OOo behave with the repacked zip file? is it faster, slower, does it need more memory when it runs? imo, changes like this should be integrated into the package build process, and sent upstream. patches welcome. same for jar files. are these extracted as fast as without your changes by the jvm? if not, then these should be left alone (and afaik there shouldn't be any jar files on the live CD). Matthias -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On Fri, 2010-10-08 at 00:07 +0800, John McCabe-Dansted wrote: > Strangely, even running advzip -z -0 > images_human.zip shrinks it by 3%, and even shrinks the corresponding > images_human.zip.gz file That's not strange, that's just entropic packing principles. You've got a bunch of assumptions that can be made about data and a bunch of compression iterations, each make assumptions about the nature of the data and some are fitting together better. I'm keen on this work since saving space allows for all sorts of goodies. Did we save space with any of the SVG cleaning or did that need to be brought up to the packaging level? Martin, -- Ubuntu-devel-discuss mailing list Ubuntu-devel-discuss@lists.ubuntu.com Modify settings or unsubscribe at: https://lists.ubuntu.com/mailman/listinfo/ubuntu-devel-discuss
Re: More LiveCD space optimizations
On Thu, Oct 7, 2010 at 10:05 AM, Louis Simard wrote: > Hey :) > > Thanks for the interest in this optimisation! Unfortunately I wasn't > pushy enough in my thread from May-June and it wasn't included in the > Maverick LiveCD. A pending question is what to do to include the > recompressed files into the archive's packages [1]. I think this will be discussed at UDS-N, see: http://archives.free.net.ph/message/20101004.065026.e553efd1.en.html > 2010-10-06 16:08 GMT John McCabe-Dansted : >> In May, Louis Simard proposed rencoding PNG files and SVG files to >> reduce their size [Quoted 1]. I note that we can save further space by: >> >> 1) Using advdef on the png files in addition to optipng. This is what >> optimizegraphics does, and this shrinks the pngs on the Maverick RC >> liveCD from about 100.1MB to 85.3MB providing a saving of 14.8MB. > > So it does; I didn't know about that. Reading the man file for advpng, > it gave a warning that it was only supported for AdvanceMAME-generated > PNG files, so I was skeptical, but it does shave off about 4% more > filesize on average with 'advpng -z4'. We could test each file to ensure the image is identical, perhaps using pngtopnm, and md5sum. This would be especially important for jpegrescan/jpgcrush, which is at version 0.0.0-1. >> 2) Recompressing gz files with advdef. Using advdef, we can shrink the >> gz files from 89.5MB to 84.8MB, and provides a saving of 4.7MB. > > That's an interesting optimisation; I didn't really know about it > either. However, I did use 7zip's Deflate compressor to recompress a > .zip file of OpenOffice.org's from 5.9 MB to 5.4 MB. The method was > rather crude, but it did the job: > > mkdir extracted > cd extracted > unzip ../file.zip > 7z a -tzip -mx=9 -mfb=258 file.repack.zip extracted/* > rm -r extracted You mean images_human.zip? I have a hunch that compressing that file wouldn't actually save space on the liveCD as I can gzip it down to 3.9MB. It may be better to leave it as an uncompressed zip, and let squashfs deal with it. Recompressing the pngs contained in the zip sounds worthwhile though. Strangely, even running advzip -z -0 images_human.zip shrinks it by 3%, and even shrinks the corresponding images_human.zip.gz file Also, there are 12MB of jar files, which are basically zip files. We can also shrink those by 5MB or so with advzip, but that doesn't seem to shrink a .tgz of them so it may not shrink the liveCD. Since zip files compress file by file, we may be able to save space on the liveCD by running "advzip -z -0" on them. That would expand them to 24MB, but reduces the size of a .tgz of them to 4.6MB, possibly saving space on the liveCD if squashfs is similarly efficient. >> 3) Recompressing jpeg files with jpegrescan. This only saves 0.5MB, >> but implementing this would add just a couple more lines of code, and >> jpegrescan does not lose any picture quality [Quoted 2]. > > jpegoptim indeed performs lossless optimisation of JPEG files by > editing Huffman tables, and it's used as the basis of jpegrescan. > However, jpegoptim doesn't make non-progressive files progressive, as > I understand jpegrescan does. This may make jpegoptim's optimisations > more transparent to applications that, for some reason, can't decode > progressive JPEGs and thus have non-progressive JPEGs in their > packages. However, most applications should be using libjpeg anyway, > so perhaps this point is moot. > >> >> Together these should shrink the liveCD by over 20MB. This is without >> even considering the .xml and .svg optimizations Louis proposed. >> >> A further 10MB could be saved by recompressing the gz files as lzma. > > At what LZMA compression level? Default (7) or --best (9)? --best Also, if we want to take replacing deflate with lzma to extremes, we could replace the deflate compression in the png files with lzma. A command that does this is "advpng -z -0 $f && lzma --best $f". I found that this could save 18.7MB. However, It may also degrade performance slightly, but I doubt it would be too significant on modern CPUs. Running unlzma on all 66MB of the .png.lzma files takes: real1m2.666s user0m6.540s sys 0m5.610s I think the user/sys are the relevant ones, and taking 12s to read every png doesn't seem too bad. The main thing is that I doubt that it would work out of the box. If we use lzma in the squashfs, just deflating them all with advpng -z -0 could reduce the liveCD size. Probably wouldn't help the installed size though. >> This seems reasonable as lzma has reasonable decompression times (e.g. >> 7ms to decompress a largish manpage like lsof). > > 7 ms? What's your CPU? :) Core2Duo E7200 @ 2.53GHz >> Since the liveCD is >> compressed anyway, it seems that if a file is compressed with gzip. it >> is worth compressing with lzma. The command "man" already seems to >> have lzma support, but we'd want to test each application to ensure >> that it functions correctly when its .gz files are replaced with lzma
Re: More LiveCD space optimizations
Hey :) Thanks for the interest in this optimisation! Unfortunately I wasn't pushy enough in my thread from May-June and it wasn't included in the Maverick LiveCD. A pending question is what to do to include the recompressed files into the archive's packages [1]. 2010-10-06 16:08 GMT John McCabe-Dansted : > In May, Louis Simard proposed rencoding PNG files and SVG files to > reduce their size [Quoted 1]. I note that we can save further space by: > > 1) Using advdef on the png files in addition to optipng. This is what > optimizegraphics does, and this shrinks the pngs on the Maverick RC > liveCD from about 100.1MB to 85.3MB providing a saving of 14.8MB. So it does; I didn't know about that. Reading the man file for advpng, it gave a warning that it was only supported for AdvanceMAME-generated PNG files, so I was skeptical, but it does shave off about 4% more filesize on average with 'advpng -z4'. > 2) Recompressing gz files with advdef. Using advdef, we can shrink the > gz files from 89.5MB to 84.8MB, and provides a saving of 4.7MB. That's an interesting optimisation; I didn't really know about it either. However, I did use 7zip's Deflate compressor to recompress a .zip file of OpenOffice.org's from 5.9 MB to 5.4 MB. The method was rather crude, but it did the job: mkdir extracted cd extracted unzip ../file.zip 7z a -tzip -mx=9 -mfb=258 file.repack.zip extracted/* rm -r extracted > 3) Recompressing jpeg files with jpegrescan. This only saves 0.5MB, > but implementing this would add just a couple more lines of code, and > jpegrescan does not lose any picture quality [Quoted 2]. jpegoptim indeed performs lossless optimisation of JPEG files by editing Huffman tables, and it's used as the basis of jpegrescan. However, jpegoptim doesn't make non-progressive files progressive, as I understand jpegrescan does. This may make jpegoptim's optimisations more transparent to applications that, for some reason, can't decode progressive JPEGs and thus have non-progressive JPEGs in their packages. However, most applications should be using libjpeg anyway, so perhaps this point is moot. > > Together these should shrink the liveCD by over 20MB. This is without > even considering the .xml and .svg optimizations Louis proposed. > > A further 10MB could be saved by recompressing the gz files as lzma. At what LZMA compression level? Default (7) or --best (9)? > This seems reasonable as lzma has reasonable decompression times (e.g. > 7ms to decompress a largish manpage like lsof). 7 ms? What's your CPU? :) > Since the liveCD is > compressed anyway, it seems that if a file is compressed with gzip. it > is worth compressing with lzma. The command "man" already seems to > have lzma support, but we'd want to test each application to ensure > that it functions correctly when its .gz files are replaced with lzma > files. We could also selectively recompress the gz files, as some .gz > files are actually smaller (by about 40 bytes) than the corresponding > lzma file. I hadn't considered this type of "transcoding" for the LiveCD. We may want to ourselves test which programs accept .lzma files in their directories in addition to .gz. Shall you do it, shall I, or shall we both do it? Also, is anyone else interested? Your point about files being compressed anyway is kind of interesting: both Deflate and LZMA recompress very poorly, so saving bytes by switching from one to the other makes sense. That would not shrink the *installed size* of these man pages much, though, because of default 4 KB blocks for ext[2-4]. > > Given that recoding SVG files can save 7MB [Quoted 1], simply recoding files > could free up 37MB for the Natty LiveCD (and presumably also reduce > the the average size of debs in the repos by about 5%). > > [Quoted 1] > http://www.mail-archive.com/ubuntu-devel-discuss@lists.ubuntu.com/msg11337.html > [Quoted 2] http://news.ycombinator.com/item?id=803839 > > I attach the script I used to check how much space would be saved. > This is purely for reproduction of my results, it is not integrated > into Louis's script. Do you want me to add to my script any of the optimisations discussed in your email? They are: Using AdvanceCOMP to recompress .png images and gzipped files; using either of jpegoptim or jpegrescan to losslessly recompress .jpg images; "transcoding" man pages from .gz to .lzma. I'm not going to add untested optimisations yet, such as transcoding *all* .gz files to .lzma. I'm still very interested in this, despite the lack of posting about the subject in the last 4 months! I've just been waiting for the guys at Debian to advise me on how to best integrate these optimisations into packages. Perhaps I should just devise a set of suitable build-depends additions (optipng, advancecomp, jpegoptim) and makefile rules for .png/.jpg/.gz, then file a single bug report on all of the packages that would benefit the most from optimisations? That way, package maintainers could opt in rather easily. - Louis [