Date: Sun, 09 Jul 2000 02:17:22 -0400 From: "Thomas M. Sommers" <tms2@mail.ptd.net> To: Brett Glass <brett@lariat.org> Cc: chat@FreeBSD.ORG, advocacy@FreeBSD.ORG Subject: Re: Emulation (Was: No port of Opera?) Message-ID: <396818F2.51CF061C@mail.ptd.net> References: <4.3.2.7.2.20000706190244.0483ad70@localhost> <4.3.2.7.2.20000706201218.04a99100@localhost> <4.3.2.7.2.20000706222258.046d9c00@localhost> <4.3.2.7.2.20000708105237.0448ca90@localhost> <4.3.2.7.2.20000708162010.050e5da0@localhost>
next in thread | previous in thread | raw e-mail | index | archive | help
Brett Glass wrote: > > At 03:06 PM 7/8/2000, Thomas M. Sommers wrote: > > > > >> 1) It may not be generally true (though IMHO it should be). > > > >If it's not generally true, then it's not an argument against Linux > >binary support. > > I should have made myself more clear. It may not be generally true > that FreeBSD users will avoid running the Linux binary under > emulation. And every one that DOES run the Linux under emulation > provides a sixfold reward to the developer for NOT doing the port: > > a) S/he saved the trouble of creating a new SKU; > b) S/he saved the expense of stocking inventory of that SKU; > c) S/he saved the non-recurring cost of engineering the port; > d) S/he can devote scarce engineering resources to a different port > (i.e. to a platform where emulation was not available); > d) S/he saved the recurring costs associated with supporting the port; > > and > > e) S/he saved the recurring costs associated with marketing the port. > > >If it's being used as an excuse, then developers will just find some > >other way to rationalize not porting to FreeBSD. > > See the six factors above. So the absence of Linux binary support will not necessarily mean more native ports, and will mean that programs that don't have native ports will not run. Which means fewer applications for FreeBSD. Which means fewer users for FreeBSD. Which means fewer native ports for FreeBSD. ... > >I think the main flaw in your argument is that even if Linux support > >does provide a reason not to port to FreeBSD, it does not follow that > >the absence of such support will cause those missing ports to be made. > > I did not say that the absence of such support would cause the ports > to be made. One still must conquer what Judge Thomas Penfield Jackson > called the "applications barrier to entry." However, without an > emulator, market forces would have their chance to work in favor of > the port as the installed base increased. Having emulation > short-circuits them. But these market forces will probably still not be strong enough to produce native ports. In a previous post you gave two reasons for considering Linux versions unsuitable: 1) lack of support, and 2) lack of performance/quality. 1) If a vendor can't or won't provide resources to support Linux versions on FreeBSD, it almost certainly can't or won't provide the much greater resources to produce a native FreeBSD port. 2) If a particular Linux binary doesn't run well enough on FreeBSD, then the vendor's decision to make a native port will be the same regardless of the existence of Linux binary support. If the binary does run well enough, it doesn't really matter that it is not native; it gets the job done. Of course a native version would be nice, but it would not be necessary. > >There are too many other factors at work, such as limited development > >resources. > > Indeed; see above. It doesn't sound as if we disagree here. We draw very different conclusions from the same facts. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-advocacy" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?396818F2.51CF061C>