Re: Note to IBM... Wuz Re: Informix Warehouse Accelerator Prerequisites

Ok Ian. I can take some time, since you insist. Writing wrong things to me is irrelevant, but to the community makes a difference. On Mon, Aug 22, 2011 at 1:18 PM, Ian Michael Gumby <im_gumby@hotmail.com>wrote: > > > So it took 20 years for this security hole to get fixed? Kinda shows how > certain product managers treat security. Oooh Snap! Sorry Jerry, I guess > closing security holes just aren't high on the customer's wish list so that > they take a back seat to a failed product. > I can easily agree with you. 20 years to fix this is too much. But again, and I wrote this to you, to this forum and on my blog: using /etc/hosts.equiv and .rhost is only insecure IF you use them for anything else beside Informix. And I really doubt someone who cares about security does it. And as usual, beside you fix something you get hammered because you don't fix it, once you fix it, you still get hammered because you took too long. it makes you wonder if it's worth doing it, and I'm glad our R&D team decided to do it. And comparing this to other "trivial" things that are being done, you easily understand that customers were not pushing to hard for this. > Ok, sorry, that was unfair. > > Really? But once written, it's out there you know... > Red's comments are valid because he is questioning what is written on IBM's > site. They say it's a prerequisite and frankly changing to ssh is *trivial*. > He's comments mentioned telnetd and how telnet itself was insecure because it's not encrypted. It missed the point that no telnetd service was needed, and the purpose of using "telnet" command. Someone else already explained this. As usual you ignore the parts that may contradict your points. > Getting back to Fernando and the Morris worm. Not everything was made > public so it wasn't captured in Wikipedia. In order to use .rhosts as an > attack vector you already need access to the machine as root. So the > question is... How did the worm get on the machine in the first place? I > think that was made public... But don't feel too bad. It tok 8-10 years for > Sun to fix it... > All that is on the net. Just don't see the point and how it connects to this. The point relevant here, is that rservices are insecure. So you'd better not use them. After that the so called "Informix issue with .rhosts" is pretty irrelevant. I just wished we had fixed this sooner because it would save a lot of useless discussions like this one. Just out of curiosity... Wall something or other is one of IBM's largest > customers as well as one of Informix's. One has to ask how many copies of > IWA have they purchased? > The reason I ask this is that there's a group from that company based in SF > looking for Hadoop expertise. Somehow even there, I doubt that Tom Deutch's > winning personality is going to get them to even consider IBM as their Big > Data vendor of choice. (Duh Winning!) > > But hey! What do I know? > > Ok, just to be serious for a second... > IWA is a dead product out of the gate. You can not make a value > proposition that makes sense. It's too expensive and doesn't scale. There > are already alternatives on the market that offer a much better value > proposition and have a referencable market base. To start with a relatively > small customer base, you are creating a further niche product. It a loss > right out o the gate. > > It's as if someone is trying to pull a rabbit out of their ass just to make > IDS relevant when customers' dollars are being spent on big data. This is a > losing strategy. This isn't to say that IDS is a bad product. It isn't. As > I have said in earlier posts, IWA is based on stuff done in the Financial > Foundation data blades. 10 years ago it would have been interesting. Today? > Not so much... > > It's a joke. Especially when the answer is staring back at them, right in > front of their face! > It takes bright product managers as well as division execs to recognize > what they have and how to properly utilize their talent. And this boys and > girls is what IBM is lacking... > Mixing IWA with big data is such a non-sense that can actually be funny. Maybe you could start by reading the papers and documentation... But for that, you'd have to allow space in your brain to something else than Hadoop* which I find it difficult, probably because "big" data also consumes a lot of brain cells... Regards -- Fernando Nunes Portugal http://informix-technology.blogspot.com My email works... but I don't check it frequently...