Date: Fri, 08 Nov 1996 08:49:43 -0800 From: "Jordan K. Hubbard" <jkh@time.cdrom.com> To: dubois@primate.wisc.edu (Paul DuBois) Cc: stable@FreeBSD.org Subject: Re: The curtain is going down on 2.1-stable in 5 days! Message-ID: <470.847471783@time.cdrom.com> In-Reply-To: Your message of "Fri, 08 Nov 1996 10:00:32 CST." <199611081600.KAA17338@night.primate.wisc.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
> I would expect that "for commericial use" means suitable for people to > *use*, not hack on. That means when someone buys a CD-ROM set and then You haven't followed this to its logical conclusion, however, and one which the evidence we've gathered so far strongly supports: If you're truly *commercial*, you'll buy whatever you need to buy in order to make it work and achieve your end-goal (which is not fiddling with FreeBSD hardware). If you were smart enough to order your machine *after* consulting our supported hardware list and/or reading the recommendations I have up at http://www.freebsd.org/handbook/hw.html, it's not even any extra work. If you order your entire PC from an established BSD PC vendor like Rod Grimes, Apache Digital, or Telsys, it's less work still - the thing comes preloaded and ready to go. In all cases, if you're commercial then you don't spend a week jumping up and down in the mailing lists about, say, IDE CDROMs, you go "oh, I should have gotten a SCSI one? Bummer. OK, back in an hour - I'm off to the store again!" In fact, it's almost become something of a problem because dealing with buggy hardware reports can be a real chore when all you're able to note on the follow-up contact is "customer has already replaced hardware and moved on." In any case, people have gotten used to the fact that a lot of PC hardware out there is crap, and buying a machine to fit known specifications is only common sense in the PC market if you're actually trying to use it for serious work. Nonetheless, taking on another of your points: > find his drive, it's not legitimate for the developers to respond (as > they frequently do): > > 1) "Oh yeah!?! Why don't YOU fix the driver, then!" > or > 2) "Get a SCSI CD drive." Both can actually be perfectly legitimate responses, depending on the situation. 1. If you've got a burning need to have something work (be you a corporation or a private user) and there's nobody able to currently devote a week to jamming away on the problem for you, then yeah, you kinda have to do it yourself. Many of our best developers and code came from people in the throes of just such motivated self- interest. Simply crying about it certainly won't fix it, and there are only so many developer hours in the day (and lots of other tasks that someone *else* is clamoring for them to finish). This is an answer you should get used to until some kind foundation decides to throw a few $mil at me so that i can hire and house some full-time, accountable, engineering talent. 2. If you're a corporation, this is good advice. Buying a new SCSI drive will cost you $200-$300, plus an hour of someone's time to go pick it up if you don't order it over the phone. Say $400 total. Having a developer struggle for just one day with an IDE CDROM will easily cost you twice this much in lost productivity - it's just not worth it. If you're a user, this advice is somewhat less easy to take and maybe you don't want to hear it, in which case at worst you're a little annoyed. On the otherhand, perhaps you were already 90% decided to get a SCSI drive now that the NEC 8X SCSI drives have dropped down below $200, and our recommendation has simply sealed your decision. You get a drive that Just Works, life is good, the suggestion was actually positive. > Note that I'm not complaining here (even though I have an IDE CD that > the install has problems with). I'm simply commenting on what seems to > be a contradictory goals: developer's paradise vs. user's system. I > hope the point is clear enough and that no flame war will ensue. There are no contradictory goals, simply less emphasis on setting impractical ones for the sake of setting them. We're a volunteer organization and things get done as people find the time and energy to do them, period. I could write down "IDE CDROMS real problem. Big problem. Get someone on this immediately!!!" in my TODO list 20 times and it wouldn't necessarily achieve anything except to use up some more bytes on my disk. I'm already doing all I can do to solve the IDE CDROM problem (along with many others), and the solution will take as long as it takes. Jordan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?470.847471783>