Magic Linux 1.1b Released
Magic Linux is a new Linux distribution, a version aimed at Chinese users. The goal is very simple: to let everyone truly use Linux, instead of localizing one Linux after another. In terms of overall architecture, Magic Linux is based on RedHat, which is an excellent foundation. In M 1.0, Magic Linux even kept the RedHat directory structure, most of the work was done through configuration files, and the SRPMS tree was also incomplete. But Magic Linux 1.1 has improved greatly and has now basically broken away from RedHat’s shadow, and the SRPMS tree is also relatively complete. However, there are still less-than-satisfactory issues such as the installation interface. Hopefully in the future we can find a way of collaborative development and task division to gradually solve these problems. From the beginning up to now, Magic Linux has roughly gone through the following stages:
0. I got tired of RedHat. I had already used redhat for more than 5 years, and could hardly find anything new in it anymore, with no freshness at all. And the release of RedHat 8.0 brought many BUGs that originally did not exist but came into being because of rh’s modifications. I saw too many Chinese-localization articles; many people localized one version and then localized the next version. There were too many such articles, and far too few articles about actual application. So I wanted to make a version, a Linux that could be used well without localization and without beautification. Of course, another goal was to make a pure, original Linux.
1. Based on RedHat 8.0, there were no major changes in large software packages (for example KDE and so on still kept rh’s versions), while some utility software was screened and some rh bugs were fixed. A great deal of the work was done through configuration files. This version was not released, because it could only be regarded as a correction of rh; it had no essential difference from the Magic localization package from a long time ago. In other words, it was only an automatic implementation of a localization solution. There was also Yangchunbaixue for the 7.x series doing this kind of thing.
2. We began fixing major software packages, such as KDE, XFree86, glibc, and so on. The combination of latest versions at that time was even newer than the current M 1.1. For example, the KDE version available then was 3.1rc5, XFree86 used 4.2.99.2CVS, and glibc used the latest glibc-2.3. So before M 1.0 was released, we had already experienced XFree86’s new mouse style and glibc’s latest threading technology. But these versions were unusable for us. Using glibc-2.3 meant giving up a large batch of software. Using X-4.2.99 meant giving up the better Chinese implementation, and the cursor in X-4.2.99 had obviously not gone through good design, being comparatively too large (although this only required modifying the xcursorgen part in src). So this avant-garde version was also not released.
3. Magic Linux 1.0 was a compromise version, with nothing especially new about it, because it was neither flashy nor outdated. Although there were some BUGs, it was already very complete, and from then on we kept developing on the M 1.0 platform. This included later compilation of KDE 3.1 and GNOME-2.2. The fact that many new utility programs and large desktop software packages, as well as XFree86 and glibc, could compile successfully shows that M’s compatibility and generality are very good. On this point, we believe it is absolutely no worse than "Tongyi Linux", because we have no ambition, and we have not made key modifications to any software just to highlight our own characteristics.
4. M 1.1. A complete version, the version we have consistently used for office work and home entertainment. The software in this version was all carefully selected, either according to everyone’s needs or based on statistics from the Internet, for example kxmleditor. Structurally speaking, this software is fairly good, but its ability to handle xml is really too weak. It also did not support Chinese. In order to adopt it, I fixed the Chinese support issue, but in the end it was still removed from the system because it was “tasteless to keep.” So the Chinese support and usability of this version are still acceptable.
For specific details about this version, please click www.linuxfans.org/nuke/modules.php?name=Forums&file=viewtopic&t=14070
Magic Linux is a new Linux distribution, a version aimed at Chinese users. The goal is very simple: to let everyone truly use Linux, instead of localizing one Linux after another. In terms of overall architecture, Magic Linux is based on RedHat, which is an excellent foundation. In M 1.0, Magic Linux even kept the RedHat directory structure, most of the work was done through configuration files, and the SRPMS tree was also incomplete. But Magic Linux 1.1 has improved greatly and has now basically broken away from RedHat’s shadow, and the SRPMS tree is also relatively complete. However, there are still less-than-satisfactory issues such as the installation interface. Hopefully in the future we can find a way of collaborative development and task division to gradually solve these problems. From the beginning up to now, Magic Linux has roughly gone through the following stages:
0. I got tired of RedHat. I had already used redhat for more than 5 years, and could hardly find anything new in it anymore, with no freshness at all. And the release of RedHat 8.0 brought many BUGs that originally did not exist but came into being because of rh’s modifications. I saw too many Chinese-localization articles; many people localized one version and then localized the next version. There were too many such articles, and far too few articles about actual application. So I wanted to make a version, a Linux that could be used well without localization and without beautification. Of course, another goal was to make a pure, original Linux.
1. Based on RedHat 8.0, there were no major changes in large software packages (for example KDE and so on still kept rh’s versions), while some utility software was screened and some rh bugs were fixed. A great deal of the work was done through configuration files. This version was not released, because it could only be regarded as a correction of rh; it had no essential difference from the Magic localization package from a long time ago. In other words, it was only an automatic implementation of a localization solution. There was also Yangchunbaixue for the 7.x series doing this kind of thing.
2. We began fixing major software packages, such as KDE, XFree86, glibc, and so on. The combination of latest versions at that time was even newer than the current M 1.1. For example, the KDE version available then was 3.1rc5, XFree86 used 4.2.99.2CVS, and glibc used the latest glibc-2.3. So before M 1.0 was released, we had already experienced XFree86’s new mouse style and glibc’s latest threading technology. But these versions were unusable for us. Using glibc-2.3 meant giving up a large batch of software. Using X-4.2.99 meant giving up the better Chinese implementation, and the cursor in X-4.2.99 had obviously not gone through good design, being comparatively too large (although this only required modifying the xcursorgen part in src). So this avant-garde version was also not released.
3. Magic Linux 1.0 was a compromise version, with nothing especially new about it, because it was neither flashy nor outdated. Although there were some BUGs, it was already very complete, and from then on we kept developing on the M 1.0 platform. This included later compilation of KDE 3.1 and GNOME-2.2. The fact that many new utility programs and large desktop software packages, as well as XFree86 and glibc, could compile successfully shows that M’s compatibility and generality are very good. On this point, we believe it is absolutely no worse than "Tongyi Linux", because we have no ambition, and we have not made key modifications to any software just to highlight our own characteristics.
4. M 1.1. A complete version, the version we have consistently used for office work and home entertainment. The software in this version was all carefully selected, either according to everyone’s needs or based on statistics from the Internet, for example kxmleditor. Structurally speaking, this software is fairly good, but its ability to handle xml is really too weak. It also did not support Chinese. In order to adopt it, I fixed the Chinese support issue, but in the end it was still removed from the system because it was “tasteless to keep.” So the Chinese support and usability of this version are still acceptable.
For specific details about this version, please click www.linuxfans.org/nuke/modules.php?name=Forums&file=viewtopic&t=14070
我完全同意设想建立DOS组织“DOS联盟” ，也就是说和Wengier、以及“起步”站长莫老师等DOS战友一起来建立这个“DOS联盟”，以发展我国自主OS（操作系统）的高度去完成我们共同的愿望。
------党委书记
------党委书记