Rayson Ho <[email protected]> writes: > On Thu, May 10, 2012 at 9:34 AM, Chi Chan <[email protected]> wrote: >> Good point, Walid. As much as so many people hate UNIVA
Who are these people? Hate is an ugly emotion, and I commend Kipling's words <http://www.kipling.org.uk/poems_pilgrimsway.htm>, preferably performed by my friends' eponymous band <http://www.pilgrims-way.net/>. Excuse the tedium of a response to the quote imaginative unprovoked personal attack. It needs a response to attempts to stymie work done for the benefit of the majority of people on this list. Community contributions mustn't be suppressed by commercial interests, and too bad that development was apparently meant to be a closed shop. > And as LSF is not > freely available, the best one in the open source world was Sun Grid > Engine It's a pity it isn't now, and will probably stay that way without a coherent community and people pitching in. > and thus I started to work with Ron and others 8 or 9 years ago > on Sun Grid Engine so that anyone could use this technology freely. I'd have thought anyone interested in improving it would be pleased that there are now many more patches contributed than "external" ones under Sun's control (and the major individual contributor seems to have been Shannon Davidson). What precisely is the problem with others doing the same thing? There's a lot of work freely available to OGS, but it seems to be necessary to suppress it rather than use it. > 2) I did not (and still don't) want to contribute to open core > software - It was unwise to contribute code under the SISSL then. People interested in the community (including Richard Stallman) typically can support such things because the community benefits from free code. Proprietary features can be re-implemented if necessary, like the binary-only OGS hwloc binding work was. > To be fair to Univa, in the Grid Engine Steering Committee > Univa only asked us twice to merge with and contrinute to their fork, > and it was Dave Love who was yelling the most and was blaming us for > not merging with Univa. Quotes? I know others didn't perceive that. It was unclear why OGS was even represented with no interest in collaboration, and being fair to Univa clearly wasn't on the agenda then. > Note that the Open Grid Scheduler is the first > Open Source Grid Engine project outside of the control of Sun & > Oracle, and I am not sure why someone forked Grid Engine again even > there was a working one (and even recognized by Oracle), then asked us > to merge with a third fork, and yet these days copy our changes into > his fork... Why is it hard to understand the explanation for doing something useful, continuing the Sun line of development, and preserving the old material? OGS didn't respond to multiple attempts to collaborate, offering contributions. Anyway, it had expunged the documentation that most of the patches were against at the time. There was clearly no interest in fixing problems that were biting me and others, like frequent or immediate daemon crashes which don't fit my idea of "working". There's only one public source of nearly the complete material from the sunsource site, including issues and revision history, but modulo an incomplete mail archive. (I was told Univa only just grabbed a second copy of material at almost literally the 11th hour, so there might only have been one copy of any of it at all.) I wonder why it was wrong to make that available with improvements. Supersetting the freely available work is clearly good for users and it shouldn't be a threat to anyone's manhood to use other people's free code. I wonder why anyone publishes "open source" if they don't want people to use it and think some sort of permission is needed in addition to the licence. (There are ~10 changes I've incorporated or followed from OGS, out of ~700 total. They are credited, even if not direct copies, unlike stuff going the other way.) > 4) Last year Dave started the "Univa FUD" thread on the day we > released Open Grid Scheduler/Grid Engine 2011.11 - note that Dave and > the Open Grid Scheduler Project did not talk offline, and it was not > us who asked to have that discussion thread started. Quite why one should consult people who attack you at any opportunity, and have nothing to do with it, I don't know. > We joined that > discussion as Dave said that it is a good case of "Libel Action" - Where? I guess I'd consider the possibility of damages put towards gridengine development from anyone who'd like to defend such an action in en English court. However it's effectively the university's integrity being attacked. > we were sorry for Dave Nice to know, but I thought it was doing OGS a favour to spell out the reasons people could have confidence in such distributions (as long as they obey the licences) when I wasn't expecting OGS self-FUDding about it. > and thus we offered Dave the Open Grid > Scheduler 2011.11 features and we planned to work together (turns out > that it did not work out - but don't blame the Open Grid Scheduler > Project - Why not? There was no response to the code I sent, with a request for help to collaborate initially on a security issue until the time I was attacked over it. I can only assume talk of collaboration wasn't sincere. > those who were CCed in the LD_* security discussions would know which > person is always in attack mode). I'd be happy to have the correspondence published. People would see that it was declared necessary for OGS customers to get fixes using (I was told) my work. However I wasn't allowed to use it myself until Oracle's embargo on the CVE raised by Debian was lifted. I expect most people would consider that unreasonable, along with not being "allowed" to communicate with others, including Debian security, about issues you've found and have proposed fixes for. It would also be clear that there was no response to the request for proposals on handling such things, let alone the agreement that's been claimed. > To be fair to both sides, I think we should just let that discussion > die off - it's all his said vs she said... Obviously irony is not dead. -- Community Grid Engine: http://arc.liv.ac.uk/SGE/ _______________________________________________ users mailing list [email protected] https://gridengine.org/mailman/listinfo/users
