Path: utzoo!utgpu!water!watmath!clyde!bellcore!faline!thumper!ulysses!andante!princeton!udel!rochester!cornell!uw-beaver!tektronix!reed!nscpdc!cvedc!markh From: markh@cvedc.UUCP Newsgroups: comp.binaries.ibm.pc.d Subject: Re: What to do about binaries Keywords: binaries binaries binaries Message-ID: <583@cvedc.UUCP> Date: 30 Jun 88 15:51:15 GMT References: <6010@megaron.arizona.edu> <8749@netsys.UUCP> <3355@bsu-cs.UUCP> Organization: Computervision, Beaverton, OR Lines: 74 Well, here comes my $.03 worth ... In article <3355@bsu-cs.UUCP>, pervect@bsu-cs.UUCP (Barrett Kreiner) writes: > (ok ladles and gentlmints, arm flamethrowers at leisure) > Compiling one and two, I say that doing away with binaries is wrong. > Splitting the binaries into their own group would be ok.. a bit of a bother, > but ok. Going to source only would be the worst thing the system could do. > (personal opinion.. ok?) I agree with you totally in this respect. For some machines, it is much easier to grab a binary than to try and find the compiler that the other guy used (which 9 times out of 10, I don't have anyway) or try to port to the one that I have readily available. Granted, you have to be aware of the posibility of a virus, but that is why I unpack them on my machine. It's non-critical anyway. > > For some ammunition for this note, I browsed comp.sources.games for > a size comparison.. Wanna to know what I found? (No. but you're gonna get > it anyway.) > > For those of you who have access to a grand and glorious SUN station, > you too can play Mahjongg, if you collect all 38 parts.. > I pulled what was available out.. the total size (up to pt. 31) is 1594k. Having written that beast, I feel I should at least defend it (some). I don't know why, it has given me more grief that benefit ;-). In an attempt to save net load, I originally sent out the "better" portion (icons) of it out as a compiled .o file. By the time I had targeted all three major platforms I would have had 8 shar files with ~ 1Mb of data. Not bad given the multiple platform problem. But the minute that I posted a .o file (nearest thing to a binary with out actually sending one) people started screaming virus! (If you aren't hiding something in there then why not send source ;-) So, after discussing it with the moderator, we decided to send out the source for the icons. So don't blame me, I did my best to keep it down. Will of the people and all that rot! (back to the discussion at hand) This is the situation you run into where you have a "standard" compiler set. People would rather see source than binaries. They (usually) have to much to lose from an attack of "virus" because they are generally serving more than one person. On platforms where the compiler set(s) range so widely, it becomes much easier to post/use a binary than to try and match the environment it was created in. (Sounds of asbestos suit being donned) I found that in general, programming practices on the smaller platforms tend to be much sloppier than that found on say a Sun or Vax, making it difficult most times to recreate the environment to compile in. > > I'm sorry, but THE largest thing in OUR archives is the kermit source files > which had an estimated transmission size of ~700k. Even jumbo-da-elephant > PICNIX had an E.T.S. of ~560k. !And! there are a whole lot more PC's running > in the world today than SUN's. > > All I'm saying is that the gripers should be looking in ALL places for > waste, not just (apparently) trying to single out comp.binaries. > Here! Here! (with the above taken into account) > Please, no direct Mail replies, because: > 1) the net SHOULD hear what you have to say to me, and > > P.S. Mr. Rose, I feel that your (apparent) re-route of followups to > /dev/null was a low blow... plus the fact it didn't work :-P Ditto! =========================================================================== Mark Holm ..tektronix!ogcvax!cvedc!exc!markh Exceptions ..sun!cvbnet!cvedc!exc!markh 126 NE Grant Phone (503)648-8307 Hillsboro, Oregon 97124 Messages only until after 6:00 PM