: The thing that makes me crazy... : Is when people casually mention that "oh, no I don't use XYZ, it's got ABC bug", that they *haven't even bothered reporting*. It drives me crazy, and should be grounds for invoking the Remote Strangulation Protocol. : -- : David Welton 2001-11-15 : http://www.advogato.org/article/378.html It drives me crazy also. -- DavidCary On the other hand I've got about two dozen bug reports in on a major Linux distro. They're approaching the two-year mark. Not one of the bugs has been solved and only one was briefly looked at before being (incorrectly) flagged as a duplicate of another bug because its description shared about three words. Where's my incentive to keep reporting bugs? -- MTR ---- As much as it is flattering to be quoted, the original is LarryWall's: : The social dynamics of the net are a direct consequence of the fact that nobody has yet developed a Remote Strangulation Protocol. ---- ''Where's the relevant RFC? And what port #, so I can program my firewall appropriately?'' Do you want the incoming or the outgoing port # ? :-) ''Does it have a fixed outgoing port # ??'' :^/ You can tunnel through :80, no problem. ;-) ---- My company's MIS department recently decided to throttle my group's network connection. Seems only fair that I should have some way to throttle ''their'' pipes. ---- Sometimes they make you jump through hoops to report a bug. It if is a simple type-and-submit form, that is fine. But if you have to register and then click 30 screens deep and learn their screwy category hierarchies just to send a simple message, then f*ck it. There may be volunteers who don't mind classifying such messages. ''I agree, but all the software I am involved in is open source, and most of it has my email or a list's address plastered all over it.'' I'm a big believer in bug reports in theory. But last time I submitted a bug report to a major open source project (not Linux, but a high profile one), I got email back telling me to f*ck off, it worked for everyone else, so therefore there was no bug and I must be an idiot. Charming. ''What letter did it start with? A, G, K, M (or F), or P? And if it started with M, is the second letter "o" or "y")'' * I '''almost''' said "nor was it Gnome nor KDE", maybe I should have. So no, cross off G and K. To really give it away: the third letter was neither q nor z. :-) ** ''Which elimanates Mozilla from the list of suspects; I'm not sure what has q as a third letter. The fact you eliminated G (though you were thinking Gnome and I was thinking GCC) eliminates a few more. Which leaves: Apache, MySQL, Postgresql, all those languages that start with P (ython, erl, hp). Not to mention XFree86, which I left out in my first guess. Any of those? No need to defend idiots, after all... lots of those have reasonable people working for them.'' ** Not XFree86, I fixed '''those''' bugs myself (and they relatively cheerfully allowed me to, 'though they were quite slow about it). I've fixed bugs in GCC (rarely), but found the reporting process to be too much of a pain to even begin to bother with (and the codebase is ugly, btw; surprising it works so well.) In general I find such high barriers to bug reporting that I usually just do quick fixes for myself, and screw reporting the fixes, which is a shame -- but once burnt, twice warned. '''This''' is a weakness of the open source movement in general; not only are bug reports a pain, even bug '''fixes''' are essentially unwelcome, whether by design or by accident. I'm sitting on multiple bug fixes for GNU '''bc''' at the moment, and have been pondering whether the (sole) author is likely to be interested or not, for months now...it just seems so much trouble to even attempt, given past experiences with other projects. ** Gawd, I'd forgotten until now: "Yggdrasil Linux is an early Linux distribution ("distro") developed by Yggdrasil, a company founded by Adam Richter. Yggdrasil Linux was described as being a "Plug-and-Play" Linux distribution, in the sense that it would automatically configure itself for your hardware, a feature that is now taken for granted." This was back in...1993 or so? Linux patch level 0.9x (0.93? 0.99?) Possibly the first commercial Linux distribution. ** Anyway, the CDROM strongly invited bug reports to "bugs@yggdrasil.com", so I sent them a long, long list of things that had cost me weeks, maybe months, of trouble, with no cursing, but lots of complaining, as if I were reporting a bug to IBM or SunMicrosystems or something (i.e. expecting the reader to be staff with no emotional involvement and a professionally disinterested, unemotional reaction). But no, it was a one-man operation, and it was as if I had emailed "adam.richter@yggradsil.com" and personally insulted him and his baby, and he sent back a multiparagraph email, foaming at the mouth and cursing at me. Ok...he must've had a bad day, these things happen. But I '''didn't know''' it was a one-man shop, else I certainly would have been more polite rather than bluntly expressing my frustrations about the bugs. I replied to say so. Never any answer, of course. These things happen, but it was traumatic. ** The author of the GutenMark utility for readably-formatting Gutenberg texts had a good attitude, I must say, which was a refreshing change (although he popped my bubble by pointing out the inappropriateness of some stuff I wanted to volunteer ;-) ** To tell you the truth, I've forgotten! It was quite a while ago. But it wasn't as famous as the above projects, it was high profile in the sense that it is used by tens of millions of people, but only behind the scenes as a library that end users don't tend to be aware exists. I remember that it was interrelated with another project for multilingual text processing, nothing else. ''At any rate, the support you get from corporate America is not frequently better; the only difference is that the tech support weasel hangs up the phone first before telling you to f*ck off.'' * True. :-) Is that RemoteStrangulationProtocol bidirectional??? The Linux kernel project itself is pretty weird; several times I submitted major subsystems (e.g. plug and play soundcard support) to the theoretically-appropriate maintainer (rather than posting to the kernel list) and was told things like "don't bother, there are political problems on that topic that won't be ironed out '''for a year or two'''." (emphasis added) ''One thing you might try: If you use a major Linux distro (Red Hat, SUSE)--and you paid for it--complain to them. Even if a useful patch doesn't make it into the main kernel tree; you might get it into the distro builds. But yeah, some of the politics in Linux-land are astounding. It's amazing that with all the utter assholes and primadonnas who work on the kernel (a fact that makes a whole lot of them think they're God's gift to the programming world), that any useful work gets done. The ability of LinusTorvalds to herd '''that''' many cats is truly awe-inspiring.'' True. Although Linus himself is quite frequently a stumbling block, despite his positive qualities that on the whole balance out his negatives. He's quite frequently quirky about what he likes and allows, technical feature-wise, versus dislikes and disallows. People have in fact claimed that certain features that were slow in coming to the Linux kernel were a proof that there was something wrong with the open source model, when in fact they were frequently simply because Linus wouldn't let the features in for some years for various reasons. Linux came along ''exactly'' at the right time to fill the gaping hole left by the BSD Unix legal problems, which weren't ironed out until Linux was snowballing like crazy. There's a bunch of us ex-BSD folks who feel slightly guilty about never getting around to going back to BSD again once it eventually became feasible (plus of course those who did, and don't feel guilty). Linus' quirkiness came close to motivating me a few times. But I can't complain too much; Linus is pretty good overall. Just not perfect, despite the news stories. I certainly wouldn't want to use the RemoteStrangulationProtocol on him, just the occasional ClueStick(tm) :-) * Though I was quite disappointed with how he handled the whole affair with Tridge, LarryMcVoy, and BitKeeper. Unless there was something that I wasn't aware of that went on... Linus's public attacks against Tridge were both uncalled-for and out of character. * ''Mostly agreed, but (a) I think there was more call for it than was obvious; I saw hints of some under-reported history on that, and (b) Linus '''is''' only human, so for him to be out of character sometimes is probably understandable.'' * The implication in some of the posts on the matter (including one from Linus) is that Tridge is among the hardcore "free software" group among Linux developers, many of whom actively despise LarryMcVoy and his business model, and might actively seek to undermine him. Tridge has been largely silent on the matter (undoubtedly on the advice of counsel, as Mc''''''Voy was threatening legal action; though from what I understand, the techniques to reverse-engineer BitKeeper are not dissimilar from those he regularly employs to reverse-engineer SmbProtocol. And nobody in the open source community complains that he might technically be in violation of a MS EULA for doing that... * Here's an interesting thought on an advantage Linux has over BSD. Due to the respective natures of the GPL and BSD licenses; porting code from BSD to Linux is perfectly legal; moving code the other way isn't. ** Excuse me for being blunt, because I don't mean to be rude, but surely that thought is so intrinsically inescapable as to be obvious "to the most casual of readers", as they say? That's simply one of the (sharply noted) tradeoffs between the GNU vs the BSD license, after all; these licensing issue tradeoffs have been commented upon to death, surely? And just BTW, some have said that that is the advantage that the BSD license has over Linux -- the value judgement of 'which is better' depends on other issues of personal and business philosophy. Myself, I'm pleased that there are a variety of open source licenses, rather than just e.g. BSD or GNU license becoming universal. The world is complex, and has complex needs. *** I know it's obvious, but it's often overlooked. And in this particular paragraph; I'm not intending to pass value judgement on either the BSD or GNU license (I happen to think both have their place). Simply pointing out that code can migrate in one direction but not the other, and that Linux is the beneficiary. (Just like the BSD networking stack reportedly wound its way into Windows at some time or another; but the reverse cannot happen--not that any self-respecting Unix would want any part of MS-originated code). How much this matters, I dunno--the Linux and BSD kernels are reportedly too different to share code without serious porting. ---- I thought I saw this protocol somewhere. Isn't it Req Choke Choke Ack ... Ack ... Ack ... Ack ''With a couple of congestion packets thrown in (ECE, CWR)? Eventually ending with FIN packet, I presume. Talk about DenialOfService!'' ---- Say, wasn't the RFC originally submitted by d.vader@evil.empire.com? And the first working implementation. ---- See ShootTheMessenger, ExtremelyInterstrangled ... AugustZeroFive CategoryWhimsy