Date: Wed, 02 Apr 1997 21:30:26 -0800 From: "Jordan K. Hubbard" <jkh@time.cdrom.com> To: Terry Lambert <terry@lambert.org> Cc: phk@critter.dk.tfs.com, ache@nagual.ru, dyson@FreeBSD.ORG, bde@zeta.org.au, current@FreeBSD.ORG Subject: Re: ufs lock panic in -current Message-ID: <19304.860045426@time.cdrom.com> In-Reply-To: Your message of "Wed, 02 Apr 1997 19:54:55 MST." <199704030254.TAA15269@phaeton.artisoft.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> Julian has my namei/nameifree and "EXCLUDE" NDINIT() op redundant > code reduction changes. > > Feel free to commit them. How can I? You've sent them only to Julian, and as far as I can see so far, he ain't talking. This entire thread *started* from my asking you to put these same changes up for FTP someplace (since you mentioned them specifically as an example of a contribution and phk followed up that he wanted in) and now you're suddenly playing The Reasonable Man over them, after engaging in another multi-mail thread about how you *weren't* going to do this. In case you were confused, I never asked you to post the complete sources for TerryOS, simply some of these changes you keep talking about. > I just sent you (posted) some Logical Name support code. Feel free > to commit that. You posted a skeleton which may be something I can flesh out into a working Logical Name facility with a bit more work invested - what benefit would there be to my commiting what you sent me just now? I'll assume you were just short of examples in this case and tossed a non-example into the pot by mistake. > You keep suggesting that I should upload my complete patches > for my working code set (which are based on a snapshot from June > of 1994). These patches were already deemed "unsuitable". I suggest you upload *some patches*, any patches, which start the ball rolling. You seem to think I have a crystal ball of some sort and know what "complete patches for your working code set" would amount to or even look like. Well I don't and I furthermore don't even *care* what your grand design looks like or if you give us only 10% or 90% of it, all I'm concerned with is you continuing to hang around making noise about how we're impeding your progress yet remaining in denial about your own issues which actually keep this from happening. It's frustrating and it's not getting us anywhere. > I respectfully suggest you take the patches that I have updated > to meet your standards of suitability, and forget about those which > you deem "unsuitable". I respectfully suggest that you be a little more careful in your choice of pronouns. I personally have rejected nothing of yours, by my standards of suitability or anyone else's. I'm the one asking you to put diffs up, along with other members of the project team, and I could personally care less how many "gotos" you use - I've never taken any particular issue with your programming style and if somebody flamed you way back when over it well then that's hardly my fault. We've all been flamed, both fairly and unfairly, and there's no such thing as a "Unified FreeBSD Opinion About Terry Lambert" if it's such a thing you're attacking. Again, you may have sent some patches to where one guy could look at them but you didn't send them to where all of -committers could. Please do that, it's nothing more less than we ask of others. > I have already shown a willingness to meet your criteria, and there > is no need for you to publically appear willing to accept patches > which you have already rejected (since my patches applied against See above. > Because you have failed to provide a set of criteria which I can apply > to a design to determine if the results of the design would meet the > criteria. Richard and others have found themselves in the same boat. Sigh. Both you and Richard want the Core Team equivalent of a McDonald's Franchise Operations Manual, complete with intructions on how long to fry the patties, what color the uniforms should be and what the minimum standards for perkiness are for the people who stand up front and talk to customers. Sorry, we just don't have that and we don't have time to write it. Furthermore, we don't work to any fixed set of criteria short of "Does everyone seem to like it? Does it smell OK? Does the author look like the kind of guy who'll support it? Can we work with him on an ongoing basis or does he have Personality Odor?" That's it. There's no design doc for the minimum and maximum amounts of delta-V allowed per release and it all comes down to *what* the changes are and *what* they buy you, hopefully expressed or implemented in code to the extent that it's possible for people who can't peer directly into Terry Lambert's brain to figure it out and reach some conclusions about it. > > You would have been working to establish bonds of trust rather > > than dictating from the pulpit, > > Bogus and argumentative. Factual and unfortunate. Address this issue and you may find those "bonds of trust" I mentioned starting to form a little more solidly, but right now, I and many of the other developers trust you about as much as we'd trust Bill Clinton at a Miss America pagent. I draw a barbed analogy in jest but I mean this in complete earnestness. The combined effect of your postings over these last 2-3 years have left me completely and utterly unable to see you as any kind of team player. > that we are "fencing" with "slippery conversational gambits", rather > than *reading* what I have *stated* are my motivations. There was > no need for you to cause this discussion to deteriorate into a public > head-bashing instead of taking it to private email, unless your Perhaps more head-bashing got introduced as I went along than was my intention at the start, and for that excess I apologise. However, I don't particularly think that a private exchange would have done much good either, to be frank. I've used stronger language with you in private and I've tried long, involved discussion and I've tried being short and to the point, brutally so at times, and I've tried just about everything else and through it all you Just Don't Get It. You think I'm just castigating you for the fun of it, perhaps, or have some deep-seated reason for treating you "differently" from the other developers, and there is absolutely nothing flawed about your own approach. Terry does not need to change, the FreeBSD core team does. Yeah, right. That's all I can say. The less polite words would not be appropriate in this forum. > the issues. Frankly, if you spent as much time in reasoned > discussion as you do honing your sword on the bones of your victims, > we would no longer have any differences to speak to: we would agree > where we agreed, and we would agree to disagree where we did not, > and that would be the end of it. It takes two to tango, pal, and I *have* tried the "reasoned discussion" approach with you too. If you're not wearing me down with huge trestises on vote-driven commit systems, whole new levels of abstration and a host of other Must Have proposals which would only overly complicate my life and totally sideline me from all my other work but would indeed somehow solve The Problem You're Having, it's generally me telling you your best chances of getting your changes into the tree and you answering how that's all Frightfully Impossible for some reason or another right now and that the mountain is just going to have to come to Mohammed this once. Until you change your MO, this is an unending topic which no amount of reasoned discussion is going to resolve. > Review the patches sent Julian. If you don't like them, tell me > why so I can correct them and you can commite them, and I will send > more patches. If you do like them, commit them, and I will send > more patches. This is the very simple formula for getting code > out of "A Lambert". I cannot judge what I do not have. I repeat, very tiredly, for what is perhaps the 3rd time: Put them up for FTP please. Thank you. I would be more than happy to judge the code over the coder if I could see the code. Jordan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19304.860045426>