Re:Convert friends - add top 10 reasons for FF her
on
Firefox 1.0 Released
·
· Score: 1
1. Security
Any really good arguments here?
My net admins won't allow it because there isn't a centralized patch server for Windows that handles Firefox/Mozilla updates. I'm not familiar enough with how MS does this, but they use SMS or SUS or something like that to download security updates and push them out to all the desktops. They don't want to depend on end-users updating clients.
I have a Kyocera 7135 and a Pitch Solo from igo.com. You can hook up any analog vga connection and ps/2 or usb keyboard and mouse. It hooks up with a USB connection. It does use a power cord, but I don't see why these things couldn't be powered by batteries. I can also get sound out of the headphone jack if I wanted to plug in speakers. I also use a really small printer over the infrared connection. The biggest limitation I have seen is the resolution of the 7135. If phones continue to get higher resolutions, it would only improve products like the Pitch Solo.
I don't know what the limit is, but I don't think it is too unreasonable to support two major versions (e.g. 7.x and 8.x) at the same time. when 9.0 hits, then it is probably time to get off of your 7.x stuff. If this is too unreasonable then maybe Redhat should consider what their upgrade schedule is for the personal stuff. All of the crabbing about having to upgrade is definately a "Proud Certificate Holder" type issue. If I found out an admin had applied no patches for a year, I would can their ass whether or not they were LPI or RHCE certified.
IT budgets would have to be mighty thin for somebody to say they couldn't afford one more box ( at least on Intel/AMD) as a test box for patches. In the course of the year, enough packages get upgraded that you don't have the base install anymore anyway. A reinstall would likely be easier ( and possibly more secure because of a better regression testing level ) than installing your old stuff and patching the hell out of it. I realize this doesn't work as well for larger single-image or exotic hardware systems, but let's face it; if you can afford those, you better have some sort of test/upgrade/rollback system in place anyway.
The biggest difference as I see it is that you probably can't legally patch a Windows machine once Microsoft says EOL. With Redhat, at least you have the source code, and very likly someone else in the same boat as you who wants a backported patch to. You have a little breathing room instead of a lot of EULA's.
What about the old winchips or cyrix chips? These were x86 clones as well and I'm sure VIA or whoever owns them now would love to fill a market like that.
SCO has continued to work with Unixware, but I haven't heard of any major releases beyond version 7. Monterey will be a different animal because of the "64bitness" and the folding in of stuff from AIX (if there are others besides IBM and SCO on this, I'm probably missing them). Monterey of course will start out at version 1. If it requires as many patches just to keep it stable and running as Unixware 7.0, then I would rather run Linux-64 until Monterey is in version 2+. My company ships alot of machines loaded with Unixware, but the main reason I think goes back to a) partnerships and b) somebody to point a finger at when something goes wrong. I think there are technical aspects to be worked on in Linux yet, but the bigger issue is the support and the scapegoat when something doesn't work. I know there are organizations providing this yet, but they seem to try to market to me. They need to direct their salespitch at my managers. If we had a good corporate scapegoat that could offer decent support, it would go a lot further in getting Linux shipped ( at least from here ).
Slashdot Mirror
1. Security Any really good arguments here? My net admins won't allow it because there isn't a centralized patch server for Windows that handles Firefox/Mozilla updates. I'm not familiar enough with how MS does this, but they use SMS or SUS or something like that to download security updates and push them out to all the desktops. They don't want to depend on end-users updating clients.
I have a Kyocera 7135 and a Pitch Solo from igo.com. You can hook up any analog vga connection and ps/2 or usb keyboard and mouse. It hooks up with a USB connection. It does use a power cord, but I don't see why these things couldn't be powered by batteries. I can also get sound out of the headphone jack if I wanted to plug in speakers. I also use a really small printer over the infrared connection. The biggest limitation I have seen is the resolution of the 7135. If phones continue to get higher resolutions, it would only improve products like the Pitch Solo.
I don't know what the limit is, but I don't think it is too unreasonable to support two major versions (e.g. 7.x and 8.x) at the same time. when 9.0 hits, then it is probably time to get off of your 7.x stuff. If this is too unreasonable then maybe Redhat should consider what their upgrade schedule is for the personal stuff. All of the crabbing about having to upgrade is definately a "Proud Certificate Holder" type issue. If I found out an admin had applied no patches for a year, I would can their ass whether or not they were LPI or RHCE certified.
IT budgets would have to be mighty thin for somebody to say they couldn't afford one more box ( at least on Intel/AMD) as a test box for patches. In the course of the year, enough packages get upgraded that you don't have the base install anymore anyway. A reinstall would likely be easier ( and possibly more secure because of a better regression testing level ) than installing your old stuff and patching the hell out of it. I realize this doesn't work as well for larger single-image or exotic hardware systems, but let's face it; if you can afford those, you better have some sort of test/upgrade/rollback system in place anyway.
The biggest difference as I see it is that you probably can't legally patch a Windows machine once Microsoft says EOL. With Redhat, at least you have the source code, and very likly someone else in the same boat as you who wants a backported patch to. You have a little breathing room instead of a lot of EULA's.
What about the old winchips or cyrix chips? These were x86 clones as well and I'm sure VIA or whoever owns them now would love to fill a market like that.
SCO has continued to work with Unixware, but I haven't heard of any major releases beyond version 7. Monterey will be a different animal because of the "64bitness" and the folding in of stuff from AIX (if there are others besides IBM and SCO on this, I'm probably missing them). Monterey of course will start out at version 1. If it requires as many patches just to keep it stable and running as Unixware 7.0, then I would rather run Linux-64 until Monterey is in version 2+. My company ships alot of machines loaded with Unixware, but the main reason I think goes back to a) partnerships and b) somebody to point a finger at when something goes wrong. I think there are technical aspects to be worked on in Linux yet, but the bigger issue is the support and the scapegoat when something doesn't work. I know there are organizations providing this yet, but they seem to try to market to me. They need to direct their salespitch at my managers. If we had a good corporate scapegoat that could offer decent support, it would go a lot further in getting Linux shipped ( at least from here ).