My musings on... a Solaris test platform

php.internals

Robert Thompson

18 years ago
Hello, I've been thinking about Solaris support for PHP and some of the difficulty surrounding reproducing bugs under different Solaris OS versions, not to mention OS patch versions. I work at a University that heavily uses the PHP code base and have personally been actively working at trying to better the code in support of Solaris quirks/bugs. Would the following be a benefit to PHP developers, or not really worth the time, or somewhere in-between? A set of systems with up-to-date OSs, logins and GDB available to PHP developers to test and debug PHP under the following OSs: - Solaris 8 x86 - Solaris 8 sparc - Solaris 9 x86 - Solaris 9 sparc - Solaris 10 x86 - Solaris 10 sparc I could try to convince the powers that be to allow me to setup this environment, if there would be interest and use. Thanks, -Rob

Brian A. Seklecki

18 years ago
On Mon, 15 Oct 2007, Rob Thompson wrote:
> Hello, > > I've been thinking about Solaris support for PHP and some of the > difficulty surrounding reproducing bugs under different Solaris OS > versions, not to mention OS patch versions.
If you cant get traction feeding Sun-specific patches upstream, you should you should work with the Pkgsrc people. Pkgsrc can save Solaris (in that Stabbing Westward sense). See: http://pkgsrc.habel.name/packages/SunOS-5.9-sparc/lang/ php-5.2.3nb2.tgz 2007-Jun-10 11:53:47 3.2M application/x-tgz php-5.2.3nb3.tgz 2007-Sep-27 16:54:00 3.2M application/x-tgz http://www.pkgsrc.org/ They're experts at resolving toolchain and portability problems. They're also not likely to turn down access to hardware for bulk-build machines. Oh and bug reports wont get closed as "Won't fix" :) ~BAS

Robert Thompson

18 years ago
Thanks Brian, I sincerely feel that many PHP developers have their hands tied in several ways. The least of which being that the reported bugs are often not immediately and easily reproducible in a sane build environment that is easily accessible to the developer. To a developer who is voluntarily tasked with fixing a bug, their entire environment is called into question or is suspect as being inexact. This same issue does not exist for the bug reporter, who may post reports with verbiage such as "look! look! it doesn't work!" and demand restitution, back-pay, pay-back and public reprimand when the often quirky, obscure, un-reproducible fix is not immediately submitted to a time-tested and stable code base. Both sides are understandable, however, it would be a benefit to both parties to have some immediately accessible environments that are up-to-date, hopelessly obscure in their own right, and offer the developer a sane benchmark with which to base a prognosis. Again, this idea may be a waste of time if this situation does not exist, but I think it's workable. -Rob Brian A. Seklecki wrote:

Nuno Lopes

18 years ago
It would be interesting, yes. My university has been discontinuing the Solaris servers and now I only have access to an ancient Solaris 7 server. Even more interesting would be to get your patches to fix those bugs :) Thanks, Nuno ----- Original Message ----- From: "Rob Thompson" <ab5602@wayne.edu> To: <internals@lists.php.net> Sent: Tuesday, October 16, 2007 3:36 AM Subject: [PHP-DEV] My musings on... a Solaris test platform

Sriram Natarajan

18 years ago
Hi I am not sure as to why do you want to look at Solaris 8 and 9 on x86 platforms. As part of the WebStack project within OpenSolaris (http://www.opensolaris.org/os/project/webstack/), we are working towards integrating and improving the PHP experience on OpenSolaris (post Solaris 10). So, you are welcome to be involved with this project and provide your valuable feedback as well. thanks Sriram. Rob Thompson wrote: