SEVERAL months ago, Glyn Moody explained why Mono and Samba are fundamentally different when it comes to patents. In other ways, the projects are similar because the older one may provide insight into the dangers of Mono.
Despite
Fedora's decision to avoid Novell's Moonlight, this software might make it into Debian repositories some time in the future. "It's clear why," says one reader. "Fedora serves Red Hat's interests and Red Hat is vulnerable to a patent attack. But Debian -- in practice -- isn't."
In correspondence with this reader we found out that in the earlier days the Samba project, developers were in a rather similar situation, in which NT was the underdog, so they were actively endorsed by Microsoft. "When NT displaced traditional Unices, this endorsement ceased, and in the end they even embraced the GPLv3," says our reader. It's important to remember that Java is currently the leader. So, Microsoft might want to 'pull a
CIFS' on it.
"I wonder if we're about to see a repetition of history? Makes me think of the historical
"Rome does not pay traitors,"" says the reader.
That's an interesting point because based on my recollection, Samba used to attend some conferences or lesser formal events where Microsoft collaborated with them. It's a bit like documentation and conferences where Microsoft, Moonlight and Mono stand together united.
Anyway, back to Samba: At a later stage, according to what Samba told the European Commission, Microsoft stopped attending conferences or inviting Samba over for collaborations. I can find the references if required, and I think it was Andrew Tridgell -- not Jeremy Allison -- who attested to this experience. Needless to say, Samba was upset at Microsoft for this. At a later stage, Jeremy Allison told an online radio show, FLOSS Weekly, that he had heard Microsoft tells its programmers to "f*ck with Samba".
Perhaps Samba was too much of a risk, being an enabler of lower-cost competition which was -- and still is -- more stable and reliable (Microsoft prefers using the word "dependable" sometimes). That was before Christmas of 2007 when things changed due to pressure from the EU (as well as Piana
et al, Eben Moglen at the SFLC and so on).
To summarise, the commonality here is that Microsoft helps some people make its protocols prevalent on rival platforms. When it becomes a de facto standard at a more universal level, then it's all about RAND and begging for information.
Microsoft wants volunteers who lead it to API domination. Then, Microsoft can knock them out of the way.
It's important to bear in mind that Microsoft was at first afraid of Mono, according to Miguel de Icaza (interview circa 2004). It didn't let them become an integral part of technical Microsoft conferences, so Mono meetings were held across the road.
Microsoft must have had its Eureka moment later. In 2008, Miguel had
his own Eureka moment, when he realised that Microsoft was sort of betraying him with those licensing entanglements Novell had agreed to. Samba was not so gullible in comparison. It protested against Novell
even before the deal with Microsoft was signed (and
immediately after). Some of Samba's developers left Novell in protest [
1,
2].
⬆
Comments
AlexH
2008-12-29 10:50:34
Documentation doesn't prevent patent infringement. The PFIF agreement has a long list of patents that Samba have to avoid, in fact.
oiaohm
2008-12-31 07:03:24
Roy Schestowitz
2008-12-31 07:52:43
AlexH
2008-12-31 09:38:22