Re: security risks

Post #13,306 by admin 10/15/01 9:59:21 AM Reply	Re: security risks It may be the only application that a user uses with a given library. It may be generic to the programmer, but not necessarily from the browser's perspective. Uh, buh? That can be the case with any library. You can say the same thing about SCGUI if you're only running a single app with it. WTF kind of a stupid argument is that?? A generic library is a generic library. It gets downloaded once and cached. You're reaching. 1. Complicate the browser design Are already coded, complete, and installed on 90+% of the desktops out there. The complexity has already been paid. Insufficient argument. 2. Increase versioning problems In what way? As opposed to SCGUI where the user has to download a completely new version of the browser when new functionality comes out or is needed, in stark contrast to downloadable JS libraries which change immediately on the server with the app that includes them? SCGUI has the versioning problems here. 3. Create security risks On a secure VPN application used via an extranet or intranet connection? In what way? 4. Increase download times Wrong. The JS is downloaded once in a library and cached. 5. More things to crash What? How so? How are there "more things to crash" than any other client-server architecture? You're reaching again. Regards, -scott anderson
Post #13,344 by tablizer 10/15/01 12:59:26 PM Reply	Skinny client, skinny needs >> Uh, buh? That can be the case with any library. You can say the same thing about SCGUI if you're only running a single app with it. << In SCGUI you don't download libraries because most of the processing is on the server, not the client. >> [Complicate the browser design] Are already coded, complete, and installed on 90+% of the desktops out there. << What do you mean by 90+%? 90+% of all desktops with browsers? So you have to emulate on the other 10 percent? >> In what way? As opposed to SCGUI where the user has to download a completely new version of the browser when new functionality comes out or is needed, in stark contrast to downloadable JS libraries which change immediately on the server with the app that includes them? SCGUI has the versioning problems here. << When it has vector support, it can draw any fricken shape or gizmo you want. However, it is designed to move most of the processing to the server. Skinny client, skinny needs. >> [Create security risks] On a secure VPN application used via an extranet or intranet connection? In what way? << In your particular case, it may not matter. However, much B-to-B is not via VPN. >> [Increase download times] Wrong. The JS is downloaded once in a library and cached. << Everytime the cache runs out, a download wait. I have seen applet downloads that take around 5 minutes on a T1 line. And, the bigger it is, the more it will need changing in many cases. >> [More things to crash] What? How so? How are there "more things to crash" than any other client-server architecture? You're reaching again. << Who said anything about client-server? ________________ oop.ismad.com
Post #13,351 by pwhysall 10/15/01 1:09:15 PM Reply	Re: Skinny client, skinny needs >> [Complicate the browser design] Are already coded, complete, and installed on 90+% of the desktops out there. << What do you mean by 90+%? 90+% of all desktops with browsers? So you have to emulate on the other 10 percent? Yes, 90% of all desktops run either Internet Explorer or Mozilla, or some other DOM/Javascript capable browser. The number of SCGUI platforms is...ummm...two. You and your buddy. That's it. Peter Shill For Hire [link\|http://www.kuro5hin.org\|There is no K5 Cabal]
Post #13,391 by tablizer 10/15/01 2:33:40 PM Reply	The "assembler" argument again After all this time I thot you guys would be able to predict my responses and save me some typing. However, you just keep proving that you don't really understand my point of view. ________________ oop.ismad.com
Post #13,395 by admin 10/15/01 2:40:19 PM Reply	Your "point of view" Is that you've come up with something new. What we are (unsuccessfully) trying to get across to you is that you are simply advocating the use of 5250/3270 to replace what 5250/3270 (at some level) evolved into. Regards, -scott anderson
Post #13,399 by tablizer 10/15/01 2:46:54 PM Reply	I dn't care whthr it's new nor not. It is better than JS+DOM ________________ oop.ismad.com
Post #13,401 by admin 10/15/01 2:56:39 PM Reply	You have yet to demonstrate that. Regards, -scott anderson
Post #14,578 by addison 10/22/01 1:06:47 PM Reply	Re: Your "point of view" What we are (unsuccessfully) trying to get across to you is that you are simply advocating the use of 5250/3270 to replace what 5250/3270 (at some level) evolved into. And ignoring the other solutions that came up to deal with the shortcomings of 3270.. (X, ISA).. (I note that Exceed now has some sort of X-plug in for Win32, so you could (if I read this right) just shove X to people with the plugin... ) I thought that sounded somewhat nifty. I still think Bryce needs to spend a couple weeks playing with X and finding out how nice it is for what he wants to do. Addison
Post #14,658 by CRConrad 10/22/01 6:47:47 PM Reply	"Plug-in"? Uhm, isn't that called an X Server...?

Welcome to IWETHEY!