On 03/04/07, Brian Cameron <[EMAIL PROTECTED]> wrote:
Kaiwai: Oh, another thing to consider. Remember that Fluendo's plugins only work with GStreamer 0.10. Unfortunately GStreamer 0.10 is *not* compatible with GStreamer 0.8, which is what we ship on Solaris 10. If there is a desire to support Fluendo's advanced plugins on Solaris 10, then it would be necessary to backport the GStreamer 0.10 engine back to Solaris 10.
Solaris 10 'classic' (the current release) as I understand it, has to weigh up the needs of those who wish for stablibility at all costs vs those who want stability but with a degree of modernity in terms of software availability. The better target would be the Solaris Express Developer Edition - which I think should also have a blurb as a 'enthuisast' and 'desktop class' Solaris platform; and target the plugins for that said platform rather than wasting time with Solaris 10 'classic'. This would be a bit of work, but not impossible, if marketing felt
it was desirable. Or, perhaps, it makes more sense to just focus on the future and make the next release of Solaris (or the Developer Express releases) the place where Solaris becomes media savvy?
Solaris Express Developer Edition needs to be, IMHO, rebranded as "Solaris Workstation Edition" - something that emphasises it as an operating system that would make a great desktop/workstation operating system. The word Express conjures up images of a 'rickety piece of rubbish', 'here is some buggy beta code, thrown into a cd, good luck!" - this then needs to be backed up with a 'respository' for people to download updates and upgrades of software so that complete iso downloads do not need to be downloaded for updates and upgrades, however, have periodic re-spins of iso images every 4months to consolidate the updates - imagine a debiab archive but using srv5 packing and pkg-get for the dependency resolving. Couple that with making the 'on' buid releases, not only as a source download, but included as prebuilt binaries in the form of packages as to allow real world beta testing by end users and developers rather than having an incredibly high entry requirement in regards to getting the ON sources compiled and installed onto an existing Solaris Express Install. With the pricing of Fluendo encoder/decoder, personally, I think that Sun should hold off until the encoder/decoder has been finalised and then go for a bulk deal, either as a one off payment or a continual payment per year for licencing - as for the business case, I think Sun management need to realise that developer interest is a direct result of desktop usage interest; those who programme with Solaris, will use Solaris as a desktop - and mind share is what Sun needs - stop looking at the direct money coming in via licence but those licences generated by software companies writing software, and developers contributing to Solaris via the fact that Solaris first enticed them in as a suitable desktop/workstation operating system. OpenSolaris needs more minds to work on the various parts, and to do that, there needs to be firstly an enticement of features which no other operating system has - multimedia out of the box for instance - be it in the form of mp3, wma, wmv encode/decoding, flash, java, and many other features. Also, the message it shares to other third parties by Sun paying out to Fluendo is this, "we are willing to work with companies, big and small" - don't see it as a money spent, see it as an investment into a third party ecosystem which is sorely lacking right now. Matthew
_______________________________________________ opensolaris-discuss mailing list [email protected]
