headers
Matthew Toseland | 30 Jul 2012 13:43
Picon

Java version issues: ideas how to fix this gordian knot?

From what I've been able to gather here (on FMS), without yet having done my own research, there are some
problems with java at the moment...

Java 6 issues.
- _32 is latest, but linux distro's are only allowed to package 26
- therefore we must support 7
- but 7 seems to use more memory and consequently crash a lot
- openjdk ditto!
- ideas?

We should probably formally move to OpenJDK, but that seems to have bug/stability problems too for some users?

_______________________________________________
Devl mailing list
Devl@...
https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
Permalink | Reply |
headers
xor | 30 Jul 2012 14:15
Picon

Re: Java version issues: ideas how to fix this gordian knot?

> We should probably formally move to OpenJDK, but that seems to have
> bug/stability problems too for some users?

My node has been running on OpenJDK for years - without any issues.
Permalink | Reply |
headers
Evan Daniel | 30 Jul 2012 18:21
Picon

Re: Java version issues: ideas how to fix this gordian knot?

On Mon, Jul 30, 2012 at 8:15 AM, xor <xor@...> wrote:
>> We should probably formally move to OpenJDK, but that seems to have
>> bug/stability problems too for some users?
>
> My node has been running on OpenJDK for years - without any issues.

Ditto.

Well, it might be more like "year" than "years" in my case. I don't
really remember. The changeover and operation since was unmemorable.

I run a higher memory limit than some will, which might be relevant.

I'm generally in favor of such a change.

Evan Daniel
Permalink | Reply |
headers
Thomas Markus | 30 Jul 2012 18:41
Picon

Re: Java version issues: ideas how to fix this gordian knot?

On 30-07-12 18:21, Evan Daniel wrote:
> On Mon, Jul 30, 2012 at 8:15 AM, xor <xor@...> wrote:
>>> We should probably formally move to OpenJDK, but that seems to have
>>> bug/stability problems too for some users?
>> My node has been running on OpenJDK for years - without any issues.
> Ditto.
>
> Well, it might be more like "year" than "years" in my case. I don't
> really remember. The changeover and operation since was unmemorable.
>
> I run a higher memory limit than some will, which might be relevant.
>
> I'm generally in favor of such a change.
No problems here running with openjdk6 / oracle-jdk7 for quite a few 
months (and I've experimented running with _very_ tight (~120mb) memory 
restrictions on jdk7).
I'm also generally in favor of such a change.

- Thomas
Permalink | Reply |

Gmane