| 1 |
On 15:38 Sun 14 Sep , Alon Bar-Lev wrote: |
| 2 |
> Hello, |
| 3 |
> |
| 4 |
> I just curious... How do you see the boostrap process for normal users? |
| 5 |
> |
| 6 |
> You need jdk to setup icedtea... So we have somekind of loop... |
| 7 |
|
| 8 |
Yes, such a bootstrap issue is hardly limited to IcedTea. |
| 9 |
The same occurs with gcc for example, and other Free Java VMs. |
| 10 |
|
| 11 |
> Users who do not want non open source package to run on their computers should: |
| 12 |
> |
| 13 |
> 1. Install gcj |
| 14 |
> 2. Install icedtea |
| 15 |
> 3. Reinstall icedtea so it compiles with icedtea and drop gcj dependency??? |
| 16 |
> 4. Uninstall gcj |
| 17 |
> |
| 18 |
|
| 19 |
You don't need stage 3 (there is no 'gcj dependency'), and stage 4 is completely optional. |
| 20 |
Other Free VMs can be used, but they also have bootstrap issues. |
| 21 |
|
| 22 |
In the long run, we probably need a binary option for lower spec. machines, |
| 23 |
as I've already said. |
| 24 |
|
| 25 |
> Portage do not support these sequences... |
| 26 |
> |
| 27 |
|
| 28 |
Well this is probably a flaw in portage then, but the solution is only |
| 29 |
a two-step emerge process: |
| 30 |
|
| 31 |
emerge gcj-jdk |
| 32 |
emerge icedtea6 |
| 33 |
|
| 34 |
I don't think that's inherently difficult. |
| 35 |
|
| 36 |
> So we have several options: |
| 37 |
> 1. Have users do this manually. |
| 38 |
> 2. Create a script similar to crossdev that perform the sequence. |
| 39 |
> 3. Add a private gcj build into the compile stage of icedtea, and drop |
| 40 |
> the jdk dependency. |
| 41 |
> |
| 42 |
|
| 43 |
I don't see any problem with 1. Why is gcj not enabled in the bootstrap |
| 44 |
gcc on the install CDs? |
| 45 |
|
| 46 |
> Any thoughts? |
| 47 |
> Alon. |
| 48 |
> |
| 49 |
> --- |
| 50 |
> |
| 51 |
> I am almost sure you already know that... But just for the record: |
| 52 |
> |
| 53 |
> * QA Notice: The following files contain executable stacks |
| 54 |
> * Files with executable stacks will not work properly (or at all!) |
| 55 |
> * on some architectures/operating systems. A bug should be filed |
| 56 |
> * at http://bugs.gentoo.org/ to make sure the file is fixed. |
| 57 |
> * For more information, see http://hardened.gentoo.org/gnu-stack.xml |
| 58 |
> * Please include this file in your report: |
| 59 |
> * /var/tmp/portage/dev-java/icedtea6-1.2/temp/scanelf-execstack.log |
| 60 |
> * RWX --- --- usr/lib/icedtea6-1.2/jre/lib/i386/server/libjvm.so |
| 61 |
> * RWX --- --- usr/lib/icedtea6-1.2/jre/lib/i386/client/libjvm.so |
| 62 |
> |
| 63 |
> |
| 64 |
> * QA Notice: Package has poor programming practices which may compile |
| 65 |
> * fine but exhibit random runtime failures. |
| 66 |
> * ../../../src/solaris/native/java/net/Inet4AddressImpl.c:385: |
| 67 |
> warning: implicit declaration of function 'gettimeofday' |
| 68 |
> * ../../../src/solaris/native/java/net/Inet6AddressImpl.c:520: |
| 69 |
> warning: implicit declaration of function 'gettimeofday' |
| 70 |
> * ../../../src/solaris/native/java/net/Inet4AddressImpl.c:385: |
| 71 |
> warning: implicit declaration of function 'gettimeofday' |
| 72 |
> * ../../../src/solaris/native/java/net/Inet6AddressImpl.c:520: |
| 73 |
> warning: implicit declaration of function 'gettimeofday' |
| 74 |
> |
| 75 |
|
| 76 |
These are OpenJDK issues that need to be raised with Sun. |
| 77 |
|
| 78 |
Cheers, |
| 79 |
-- |
| 80 |
Andrew :) |
| 81 |
|
| 82 |
Support Free Java! |
| 83 |
Contribute to GNU Classpath and the OpenJDK |
| 84 |
http://www.gnu.org/software/classpath |
| 85 |
http://openjdk.java.net |
| 86 |
PGP Key: 94EFD9D8 (http://subkeys.pgp.net) |
| 87 |
Fingerprint = F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8 |
Archives