仕事用に新しいノートPCが届いたので、Fedora Atomic Workstation(Fedora Silverblue)をインストールした。
いろいろ躓いたのでメモ。
一番大きなプチフリ問題から。
とにかく一瞬固まる。しかも頻繁に。
問題はこれらしい
Bug 1509294 - Random freezes with nouveau driver on Lenovo Thinkpad P50
対策としてnouveauをblacklist送りにしてintel driverを使った。
今の所平気ぽい。
これ絡みでgrub.cfgを変更しようとしたら/etc/default/grubは見ないらしい。ここをいじっても反映されない。
変えるにはrpm-ostree ex kargsコマンドを使うらしい。
https://github.com/projectatomic/rpm-ostree/pull/1013
最後にSELinux絡み。
ホストになんでもぶっこんで使おうとするといちいちrebootしないといけなくてめんどくさいのでcontainer環境を用意しないといけない。でもhomeをマウントして使いまわそうとするとSELinuxで躓く。permissiveで使うのがよさそう。
June 22, 2018
December 1, 2017
i18n FAD Pune 2017
I've joined i18n FAD, Fedora Activity Day in Pune last week, 20th to 22nd November. let me share you all my experience during FAD.
First day, In the morning, I had some presentation for Fonts and Rendering in Fedora, particularly spoke about what we can propose for f28, what we are facing issues around them to share and then try to address in this event as much as possible.
After me, listened on others' talks, Pango contribution experience from Peng Wu, IBus 1.6 plans and emoji demo from Takao Fujiwara-san, ibus-typing-booster and emoji-picker demo from Mike Fabian. also glibc locale stuff from Mike and Rafal Luzynski. nominative vs genitive talk was quite interesting.
After some lunch break, Transtats update from Sundeep Anand and CLI tool demo from Parag Nemade, and Automated Testing and demo from Pravin Satpute and Pooja Yadav. the automated testing for fonts what Pravin proposed is to use hb-shape and compare the result of the analysis. I guess this testing method may works for some purpose but not for some. on the other hand, what Pooja proposed is to compare the result of the actual rendering with OpenCV which is quite opposed approach to that. well, that is to test the input method actually but could be used for rendering too.
Also f27 i18n test day report from Satyabrata Maitra, Bhushan Barve and Sandeep Shedmake. and then Web testing from Vishal.
Second day, I've taken a look at Font SIG stuff and picked up font packages missing the font information at wiki and posted it at the list. I got a lot of feedbacks on it. thanks all.
Aside from that, I've joined discussions about automated testing for fonts and langpacks. even though those topics were discussed in third day too, we didn't have any conclusion yet or still ongoing. so we need to keep working on it.
Third day, I have helped Parag who submitted a PR to improve the spec file template for font packages. reviewed and had a comment on it. thanks Parag for his effort. the templates has been up-to-date and improved a lot.
Also proposed an update against it to FPC and updated wiki too.
For my impression, I think we had more detailed discussions than where we had last time and that was also the purpose of this event. in that sense, we could say the event got an success and should helped a lot for our next step.
Thank you Parag and Sundeep esspecially for organizing this event! I had very nice experience during visiting Pune because of their helps.
First day, In the morning, I had some presentation for Fonts and Rendering in Fedora, particularly spoke about what we can propose for f28, what we are facing issues around them to share and then try to address in this event as much as possible.
After me, listened on others' talks, Pango contribution experience from Peng Wu, IBus 1.6 plans and emoji demo from Takao Fujiwara-san, ibus-typing-booster and emoji-picker demo from Mike Fabian. also glibc locale stuff from Mike and Rafal Luzynski. nominative vs genitive talk was quite interesting.
After some lunch break, Transtats update from Sundeep Anand and CLI tool demo from Parag Nemade, and Automated Testing and demo from Pravin Satpute and Pooja Yadav. the automated testing for fonts what Pravin proposed is to use hb-shape and compare the result of the analysis. I guess this testing method may works for some purpose but not for some. on the other hand, what Pooja proposed is to compare the result of the actual rendering with OpenCV which is quite opposed approach to that. well, that is to test the input method actually but could be used for rendering too.
Also f27 i18n test day report from Satyabrata Maitra, Bhushan Barve and Sandeep Shedmake. and then Web testing from Vishal.
Second day, I've taken a look at Font SIG stuff and picked up font packages missing the font information at wiki and posted it at the list. I got a lot of feedbacks on it. thanks all.
Aside from that, I've joined discussions about automated testing for fonts and langpacks. even though those topics were discussed in third day too, we didn't have any conclusion yet or still ongoing. so we need to keep working on it.
Third day, I have helped Parag who submitted a PR to improve the spec file template for font packages. reviewed and had a comment on it. thanks Parag for his effort. the templates has been up-to-date and improved a lot.
Also proposed an update against it to FPC and updated wiki too.
For my impression, I think we had more detailed discussions than where we had last time and that was also the purpose of this event. in that sense, we could say the event got an success and should helped a lot for our next step.
Thank you Parag and Sundeep esspecially for organizing this event! I had very nice experience during visiting Pune because of their helps.
December 13, 2012
Goodbye, GitHub
According to their decision that has been posted at their blog, I have also decided to move my projects to bitbucket.org, which had been hosted at GitHub. bitbucket.org has sufficient features to replace it and provide the upload space too!
They are encouraging the users to use downloadable source code archives feature, but it doesn't help for the projects that requires autotools and any bootstrap process, because bootstrapped-files won't be usually checked into the repository and doing so is a bad idea. speaking of autotools, they are checking the generated version and runtime version. it's easy to make the downloaded archives unusable.
Anyway, I'm now happy with bitbucket.org.
They are encouraging the users to use downloadable source code archives feature, but it doesn't help for the projects that requires autotools and any bootstrap process, because bootstrapped-files won't be usually checked into the repository and doing so is a bad idea. speaking of autotools, they are checking the generated version and runtime version. it's easy to make the downloaded archives unusable.
Anyway, I'm now happy with bitbucket.org.
November 25, 2012
Bits for cygwin + mintty + MSVC configuration
所用でわりとまともに動くcygwin + MSVC環境が必要になったので、将来のハマりどころ回避のためにメモ。
November 24, 2012
July 23, 2012
libeasyfc 0.9
Well, this article is just to advertise my efforts on libeasyfc again. libeasyfc is a library to allow you to add/modify/remove the fontconfig configuration for your preferred fonts.
Currently it has the following features:
- Set an alias for sans-serif, serif, monospace, cursive and fantasy per languages.
- Enable/Disable hinting, auto-hinting, embedded bitmap and anti-aliasing.
- Set/Unset the sub-pixel geometry and the hint style for auto-hinting.
No GUI in this project, but CLI is available as the example code. or you can use fonts-tweak-tool for GUI. modified version for new release of libeasyfc is also available in my github.
Enjoy!
June 25, 2012
Fontconfig in scheme
This is just experimental and wrote it for learning purpose or testing because it may be helpful to consider modifying the data structure for the better matching rules in the future.
The code just tries to build the pattern from the configuration as similar as possible, but not trying later process. I mean to match the best font in the system against it, like FcFontMatch() does. so strictly speaking this doesn't work as the language binding in scheme but if anyone wants to try, you could use it perhaps.
The code snippets are available here.
just to share this with you.
February 12, 2012
November 23, 2011
SPPUSH for CyanogenMod
November 16, 2011
Rendering on Emacs stopped working on GNOME3
It happens after "suspending" from Emacs with C-z or evaluating (suspend-frame) or whatever. though it still accepts inputs from the keyboard. and still works fine on KDE say. so the problem may be in GNOME perhaps.
If you are facing unfortunately, there are the way to be recovered from this situation; try C-x 5 2 to make an frame. you will get back the well-rendered window there and C-x 5 1 on new frame to get rid of the broken window. it saves you and your modified files.
To avoid this situation forever, I have decided to have an advice for suspend-frame in my .emacs as a workaround like this:
If you are facing unfortunately, there are the way to be recovered from this situation; try C-x 5 2 to make an frame. you will get back the well-rendered window there and C-x 5 1 on new frame to get rid of the broken window. it saves you and your modified files.
To avoid this situation forever, I have decided to have an advice for suspend-frame in my .emacs as a workaround like this:
(defadvice suspend-frame (around my-suspend-frame () activate)
(if (y-or-n-p "frame is gonna iconify ")
ad-do-it))
Enjoy,
May 12, 2011
PES-6600
他にも候補はあったけど、決め手は高輝度白色LEDが付いていたこと。試しにつけてみたけど、結構明るい。LED側のスイッチのON/OFFで点灯と点滅を交互に切り替えられる。暗い夜道を歩くなど、有事の際には役に立つかもしれない。
April 20, 2011
March 30, 2011
March 20, 2011
January 6, 2011
im-chooser 1.4.1 released
There aren't so much changes at all. this release supports imsettings-1.0.0.
For more information about im-chooser, visit here.
For more information about im-chooser, visit here.
January 5, 2011
imsettings 1.0.0 released
3 years ago, this project has been branched from im-chooser, GUI frontend to select the input method for desktop use. there has been some dozens of bugs -- even in this release probably but anyway ;) -- being fixed in releases.
Today I'm pleased to announce this release finally, with the support of the majority toolkits/desktops in the opensource world.
Today I'm pleased to announce this release finally, with the support of the majority toolkits/desktops in the opensource world.
December 21, 2010
Demo: rendering PostScript on Hieroglyph
It's been a while to post something on this blog. and I'm demonstrating this second times, but with the different code-base.
I don't have too much words to talk about this yet; This is too simple and too slow, plus may be too bored to have a look. but it should be a big step toward seeing what Hieroglyph is. anyway here is a demo video:
Note: just uploaded the original ogg video file here.
I don't have too much words to talk about this yet; This is too simple and too slow, plus may be too bored to have a look. but it should be a big step toward seeing what Hieroglyph is. anyway here is a demo video:
Note: just uploaded the original ogg video file here.
June 25, 2010
A hack to allow editing the charset and the lang in a font on fontconfig
Background
fontconfig is scanning a font and gathering the information to see what kind of character sets it's possible to support. being used to pick up the certain font according to a request from a user. having said that all of the fonts doesn't necessarily have certain coverage of the glyphs for certain character sets. so given that missing glyphs are not a big deal or still looks good with a fallback font for them, there should be needs to use such fonts for the certain language anyway.
Syntax
After this hack, two elements are available. <charset> to update the character sets information a.k.a. FcCharSet and <langset> to update the language coverage information a.k.a. FcLangSet. in either case, the element has to have the language name that is available as orth file in fontconfig but without .orth suffix, as CDATA. it looks like:
And you have to run fc-cache again to re-generate the fontconfig cache file because the above rule only affects the scan phase. then starting to match HanaMin for serif:lang=ja in this case. which originally has zh-hk, zh-mo and zh-tw coverage.
Code
See private git repo
fontconfig is scanning a font and gathering the information to see what kind of character sets it's possible to support. being used to pick up the certain font according to a request from a user. having said that all of the fonts doesn't necessarily have certain coverage of the glyphs for certain character sets. so given that missing glyphs are not a big deal or still looks good with a fallback font for them, there should be needs to use such fonts for the certain language anyway.
Syntax
After this hack, two elements are available. <charset> to update the character sets information a.k.a. FcCharSet and <langset> to update the language coverage information a.k.a. FcLangSet. in either case, the element has to have the language name that is available as orth file in fontconfig but without .orth suffix, as CDATA. it looks like:
<match target="scan"> <test name="family"> <string>HanaMin</string> </test> <edit name="charset" mode="assign" binding="strong"> <charset>ja</charset> </edit> <edit name="lang" mode="assign" binding="strong"> <langset>ja</langset> </edit> </match>
And you have to run fc-cache again to re-generate the fontconfig cache file because the above rule only affects the scan phase. then starting to match HanaMin for serif:lang=ja in this case. which originally has zh-hk, zh-mo and zh-tw coverage.
Code
See private git repo
June 1, 2010
How to reset the background on GNOME desktop (with unusual way...)
ニッチな需要かもだけど、どうなってんのという感じで。
背景画像や背景色の変更はGNOME上ではgnome-appearance-propertiesで行う。ここを一度でも変えると個々人の設定として記録される。ディストリビューションによっては毎リリースごとテーマを変えてくるものもあり、新しいバージョン入れた!って気分を味わえなくてつまらないので、これを強制的にリセットする方法を探してみた。
といってもそんな大した話ではないけど、まずひとつ。/desktop/gnome/background/picture_filenameというキーがあって、こいつをクリアすると背景画像がデフォルトのものに戻る。が、しかし、再びgnome-appearance-propertiesを起動すると設定したものへまた戻ってしまう。
で、もうひとつ。$HOME/.gnome2/backgrounds.xmlというファイル。ここにも背景画像、背景色に関する情報が記録される。両方消すとデフォルトのものが正しく適用されるようになる。
おしまい。
背景画像や背景色の変更はGNOME上ではgnome-appearance-propertiesで行う。ここを一度でも変えると個々人の設定として記録される。ディストリビューションによっては毎リリースごとテーマを変えてくるものもあり、新しいバージョン入れた!って気分を味わえなくてつまらないので、これを強制的にリセットする方法を探してみた。
といってもそんな大した話ではないけど、まずひとつ。/desktop/gnome/background/picture_filenameというキーがあって、こいつをクリアすると背景画像がデフォルトのものに戻る。が、しかし、再びgnome-appearance-propertiesを起動すると設定したものへまた戻ってしまう。
で、もうひとつ。$HOME/.gnome2/backgrounds.xmlというファイル。ここにも背景画像、背景色に関する情報が記録される。両方消すとデフォルトのものが正しく適用されるようになる。
おしまい。
May 20, 2010
Fedora yum repository for Mozc available
I've just created yum repository metadata for Mozc on my fedorapeople space. you can install the latest package after installing mozc.repo file under /etc/yum.repo.d on your system.
Currently it contains the binary rpms for i686 and x86_64 only on Fedora 12 and 13. but not for ppc nor ppc64 because unfortunately the build failed.
For another news, started to include mozc-config in the package. though I pondered to make another package for that, it was difficult to do so because it requires the build environment of Mozc for header files and static libraries and objects.
Enjoy,
[Update: 2010/05/28]: fixed a fail on i386 box to fetch the metadata via yum; sorry for inconvenience.
[Update: 2010/06/22]: fixed the repo file and URL because it has been changed due to the restructure of the package.
Currently it contains the binary rpms for i686 and x86_64 only on Fedora 12 and 13. but not for ppc nor ppc64 because unfortunately the build failed.
For another news, started to include mozc-config in the package. though I pondered to make another package for that, it was difficult to do so because it requires the build environment of Mozc for header files and static libraries and objects.
Enjoy,
[Update: 2010/05/28]: fixed a fail on i386 box to fetch the metadata via yum; sorry for inconvenience.
[Update: 2010/06/22]: fixed the repo file and URL because it has been changed due to the restructure of the package.
Subscribe to:
Posts (Atom)