Plymouth on Fedora 11

July 3rd, 2009

Plymouth has been working perfectly on my Fedora 10 box after I edited my grub.conf file to match the vga modes required for my display. After upgrade (or fresh install) to Fedora 11, it stopped working again. No idea, what went wrong. Tried digging through the problem and came across the nice guide: http://www.my-guides.net/en/content/view/161/26/2/13/

All I needed to do apart from modifying grub.com was, change the theme to solar from default (text).

su – root
plymouth-set-default-theme solar
/usr/libexec/plymouth/plymouth-update-initrd

Thanks my-guides.net

Posted in Fedora | No Comments »

Turning off: http://l10nblogs.indianoss.org/

July 2nd, 2009

Last weekend, I setup my blog with WordPress installed on a new domain (www.ankit644.com) and migrated my blog posts, comments everything from http://l10nblogs.indianoss.org/ properly. It seems to be working fine now.

So, I will be taking down http://l10nblogs.indianoss.org/ by next weekend…

Posted in Miscellaneous | No Comments »

How can you fail spammers attacking your website?

June 28th, 2009

Do you think we can stop spammers attacking our websites?

I think, NO! There is no way to stop spammers trying to attack your website, especially when your website starts getting popularity. But there are certain ways to prevent your website from spam attacks. Captcha is a way to fail spammers attacking your website.

Some rubygems are available for Captcha implementation under Ruby On Rails. You can find all of these Captcha rubygems here. Simple Captcha by Expressica.com is very popular Captcha rubygem out of the mentioned. A simple captcha Ruby/CAPTCHA is also good. ruby-recaptcha is another good example. Apart from all these, captchator.com also does provide a very good facility of using captchas in your website, which never requires you to install anything on your server to implement captcha validation.

But, the one I liked best suitable to my needs is Turing. There are number of reasons, why I liked it:

  • Easy to install (e.g. gem install turing)
  • Easy to configure (e.g. add “require ‘turing’” to config/environment.rb file)
  • Easy to implement (check here: example, entire manual can be found here)
  • Inbuilt glossary of words, which can be modified to meet individual’s needs, used to pick the random word for captcha images.
  • Various images used to pick as random background image for captcha
  • And the best part is: automatic removal of captcha images when you validate user input text with image text. (e.g. tc.valid_answer?(id, text))
  • No dependency on any website for captcha images or validation. Everything is done locally on the server
  • Turing is Licensed under GPLv2

There is no active development on Turing since 2007, however it fulfills most of my needs and works nicely without any kind of problem.

Despite these good features, there are improvement areas:

  • Along with making captcha images complex for machines to read, it should be made easy for humans to read. :)
  • OCRs might be able to decode your image text even though the captcha images look distorted.
  • Auto-removal of captcha images on the validation, which is being taken care by Turing, but I guess not by others.
  • Auto-removal of captcha images on the page unload event. Imagine a spam script continuously reloads a page every second. Your space on the server might get filled up in minutes.
  • Accessibility featues for disabled. Some captchas like ruby-recaptcha is successful in this area upto some extent.

I am still trying to find better ways of auto-removal of captcha images on the page unload event. Let me know, if you have got some inputs on it with respect to Turing rubygem.

/Ankit

Posted in Ruby On Rails | 1 Comment »

Gujarati locale and standards…

June 15th, 2009

Few years before I have got a chance to modify the Gujarati locale file, which was initially created by Karunakar. Since the locale file plays a very important role in computer localization for any language, it was really important task for me. So, I had to make sure that any kind of change I make to this file is very accurate. After taking reference from Dictionaries I had entered the appropriate entries in the locale file.

Since last couple of days, there has been a discussion happening on Indlinux-group mailing list for issues posted against Gujarati locale. Issues were basically posted by French guy called Bernard, who can speak fluent Hindi and can read Gujarati to some extent. I felt really good when I saw the initial posting, since I found someone apart from the existing localization contributors, interested in Gujarati localization.

The major issues highlighted in the discussions were:
1. Rupee symbol in Gujarati locale
=> According to Bernard and few others as well, Rupee symbol should be “રૂ” rather than “રુ” or “૱”.
=> I have done my secondary, higher secondary education in pure Gujarati medium and used or seen “૱” only.
=> Unicode chart also points to the rupee symbol for Gujarati as “૱” (U+0AF1)
=> have taken a reference from SSC (10th) Maths Question Book from Gujarat Secondary and Higher Secondary Education Board website: http://www.gseb.org/gseb/download/student/question-bank/gujarati-medium/std-10/Maths.pdf – Page no. 12, 13. Maths_PDF.p65
2. Translation of Thursday into Gujarati
=> Thursday translates to “ગુરુવાર” in Gujarati. But here the spelling should be “ગુરૂવાર” according to Bernard.
=> There is no fault from Bernard’s side since the calendar that he sees at home displays “ગુરૂવાર”.
=> In fact when I was editing the Gujarati locale file initially, I also assumed that the spelling should be “ગુરૂવાર”. But I had to refer to the standards, dictionaries, etc. and I realized that the correct spelling is “ગુરુવાર”.

Last weekend I got a chance to visit my hometown Bilimora. While reading newspaper (Gujarat Samachar), I have noticed they have also been using wrong symbol for Rupee (રૂ|.) and Thursday ( ગુરૂવાર ). Same wrong spelling for Thursday I found in a calendar at my house from Gujarat Samachar. You can see that here:

Wrong_Calendar

There was another calendar in my bedroom :)

Correct_Calendar

So, I think the print media has been using wrong spellings for some words, which creates a confusion among users and they tend to use wrong words which they have been seeing often.

I am not a Gujarati linguist, but I have done my school study from Gujarati medium and know Gujarati upto well extent. I have modified the locale file according to the definitions provided in the Dictionaries and standards. However, this discussion needs an authentic verification from a language expert.

Thank you for reading!
Ankit

Posted in Gujarati, Localization | 2 Comments »

Mercurial (On Linux) for Mozilla Localization

September 12th, 2008

Mozilla localizers have been using CVS since many years for contributing on Mozilla (Firefox/Thunderbird) localizations. Recently, Mozilla changed the VCS (Version Control System) from CVS to Mercurial (Hg). I never worked on Hg anytime before, had some confusions setting up the username for submitting my patches to Hg system. After spending some hours, figured out some easy steps to configure username. I have tried to put the basic and simple steps required for the Mozilla Localizers to get their work started on Hg VCS…

  • Installation
    Fedora, OpenSuse

    yum install mercurial

    Debian

    apt-get install mercurial

    Gentoo

    emerge mercurial

  • Configuration
    ~/.ssh/config

    Host hg.mozilla.org
    User user@host.domain

    (your-local-hg-root)/.hg/hgrc

    [paths]
    default = http://hg.mozilla.org/l10n-central/ab-CD
    default-push = ssh://hg.mozilla.org/l10n-central/ab-CD

    Setting up a username
    Various ways to setup username, which can be used to display as an author for submitting…

    1. If you specify a -u option to the “hg commit” command on the command line, followed by a username, this is always given the highest precedence.

      hg commit -m “commit message” -u “Firstname Lastname <email@domain>”

    2. If you have set the HGUSER environment variable, this is checked next.

      export HGUSER=”Firstname Lastname <email@domain>”

    3. If you create a file in your home directory called .hgrc, with a username entry, that will be used next. The contents of this file should look like, as shown below:

      # This is a Mercurial configuration file.
      [ui]
      username = Firstname Lastname <email@domain>

      As this configuration is going to be common across various Mercurial repositories on your machine, you can make Mozilla specific settings of the username in the (your-local-hg-root)/.hg/hgrc file by adding the contents exactly in the same format as shown above.

    4. If you have set the EMAIL environment variable, this will be used next.

      export EMAIL=”<email@domain>”

    5. Mercurial will query your system to find out your local user name and host name, and construct a username from these components. Since this often results in a username that is not very useful, it will print a warning if it has to do this.

    The name and e-mail Id configured will be displayed as author on the actual Hg repository: http://hg.mozilla.org/l10n-central/ . It doesn’t matter which e-mail ID is used as an account to access the Mozilla repository.

  • Download L10n

    hg clone http://hg.mozilla.org/l10n-central/ab-CD/

  • Submit L10n

    hg commit -m “message identifying the change (bug 555)”
    hg push

  • References:
    http://developer.mozilla.org/En/L10n_on_Mercurial
    http://hgbook.red-bean.com/

Posted in Firefox, Linux, Mercurial | 2 Comments »

« Older Entries
Newer Entries »