I suspect that I don't really have a PRACTICAL choice, here.
The XP option has long set sail -- I'm *sure* MS is no longer making XP licenses available (to non-profits or ANYONE!).
And, the normal update cycle of 7even is determinedly moving those boxes to the W10 model. At some point, one can expect
7even to simply *become* W10. Installing a "tuned" version of 7even (selectively NOT installing updates that compromise it) and then locking it down (as I did with XP) will leave those boxes as "vulnerable"/compromised as running XP in 2016 would.
[No desire to manually update individual machines on an ongoing basis. And, not practical to do all the bookkeeping to know which machines need which updates, etc. Nor expecting kids to make their machines available to me -- while expecting me to be finished lickety-split! I'd have to set up an online service to do this (and probably run afoul of MS's legal]
I used to deal directly with Peter Norton back in the Norton Utilities days. If we had a problem, I got on the horn with Peter and the problem was solved. Symantec Support is a total oxymoron. Symantec end-point is just a way to turn a fast computer into a doorstop- particularly in a network situation connected to the internet.
I haven't done any coding for the last 7 years but I built the original e-commerce site for the insurance company. It worked very well, but it wasn't fancy. We hired a developer to take over that work so I could concentrate on the other work that needed to be done. She coded everything by hand but was unable to produce some of the features management determined they needed, so we moved to a new developer who uses some more advanced programming tools and produces webs that work on mobile devices as well - since many of our clients depend almost exclusively on their mobile devices to access the net.
By "your" I was referring to the developers that build that type of web - not "yours" in particular.
You don't know what I know about webpage coding. I know I'm not up to date on the latest methods - but I've done a fair bit of web development in years past
Most of the time compatability mode is not required. When a site doesn't work, every one I have run into runs well on compatability mode. Most that work fine without will give problems if compatability mode is used. Most of what causes problems for us falls into the class known as "portals"
Most of the time you are right, Firefox works. BUT Firefox does not integrate with the major insurance application that is used heavily in the office - IE does - as does Outlook.. Using the same browser for everything is a lot more efficient than using different browsers for different applications.
Yup. Each time I've looked to Linux for an answer, I've come away wondering if I shouldn't, instead, be asking a different QUESTION! :<
Ditto. In my case, my vocation *and* avocations lie with this sort of technology. So, I can make it do all sorts of things for me (or my clients).
But, I'm not keen on maintaining something that shouldn't need maintenance. When was the last time you upgraded the software in your microwave oven? Dishwasher? Automobile?? Did any of those things suddenly become LESS of their former selves because of the "missed" upgrade?
I worked with a guy who was constantly tweaking our "math library" (little "subprograms" that we employed in our products to perform certain calculations). No idea what he hoped to achieve with all the effort expended on that task (is he expecting to improve on 2+2=4? Is there a MORE CORRECT answer??).
We would jokingly refer to his efforts as "RE-bugging the math library" (cuz his changes inevitably broke SOMETHING)
Why do I want to move to 7even (or 8 or vista or...) when those other apps are working fine under XP? Moving all the other apps brings up licensing issues (hardware locked) for them, risks compatibility problems, means the new machine has to support the old interfaces (e.g., serial and parallel ports, PCI-X slots), etc.
Instead, move *just* what "needs" to be moved. An extra machine is a small price to pay for NOT having to migrate all the WORKING and PROPERLY CONFIGURED software!
Actually a LOT of Voip phones here in Canada - Rogers internet also provides Rogers HomePhone - which is VOIP and is fairly heavily subscribed. A fair amount of MagicJack too
Fibernetics and Newt have a pretty substantial business VOIP presence here in the Golden Triangle area as well.
Buffer overrun exploits? C'mon, you see this problem *once* and you put mechanisms in place to ensure it never appears again, in any newly developed code.
Unless, of course, you've got lousy development practices!
Why do file sizes get reported differently depending on the "viewport" through which you are examining them? Isn't this a problem that you solve *once* and then reuse? Or, do you let each developer come up with his/her own notion of how to report a file size?
Why can I create paths that are too long to actually traverse? Isn't there ONE set of routines for manipulating these "objects"?
Didn't they "learn" that startup macros were A Bad Thing in the EARLY days of MSWord/Excel/etc.? So, why create the exact same mechanism with autorun? Then, require you to edit registry entries to "fix" the problem (again) -- that THEY had introduced?
It's clients systems that have me redigging ditches. I have reduced my workload since I no longer have the responsibility of maintaining the servers and maintaining the backup, but just keeping up with the other changes keeps me busy a couple days of the week.
I still use a lot of old toold too - being forced to upgrade some that don't work on 64 bit OS
There are 1600 attempted connections to port 3544 which is allegedly used in the "customer experience program" (spyware that you possibly can opt out of -- one has to wonder why MS would choose to enable this, by default?)
There are ~1200 attempted connections to port 443 (HTTPS) and ~400 to port 80 (HTTP).
Another 600 attempts to port 53 (DNS) and another 600 to
137 (NetBIOS). These can be simple network discovery probes. Or, "calls out" prior to initiating further connection attempts to other "named hosts".
The fact that so many were (apparently) hard-coded (port 53 is blocked so where could those IP addr's have come from if not encoded in the binary?) is interesting/suspicious. Of course, there may be some uncertainty as to when the ports were blocked; any name resolution that occurred early in the installation may have been cached before the firewall was erected (MS may *require* a live connection for the install to work?)
[This is why having DETAILED logs of the installation process are important: when did you do *each* action. So, you can later audit YOUR actions to see where discrepancies may have crept in.]
By my count, we're at 4000 connection attempts (all blocked) on a machine that has "not used the Windows 10 installation at all" (the author was asleep during the test). After having "disabled three pages of tracking options"
The IP addresses involved resolve to msn.com and akamaitechnologies.com (akamaitechnologies is one of those mechanisms that allows you to be tracked across HTTP domains)
The use of these well known ports (80/53/443/137) may be innocent. Or, it may be a surreptitious attempt to probe *through* external firewalls (cuz those ports tend to be NEEDED to be open for their NORMAL, intended traffic so one can exploit them to route specific data to external hosts tunneling through them!)
When our support people come looking for help when a client complains about slow operation, the first question I ask is "Is the AV configured to exclude all our directories?"
Then we move on the the network mayhem. I recently found a site that had configured the workstations on their LAN to use 8.8.8.8 for DNS.
Oh, you wanted to connect to anything but a modem? There's this company in Australia called Trumpet run by a guy called Peter Tattam that has this thing called Winsock...
There's hope though. I have 4 desktops on my Linux box at home, 8 at work. I understand Win10 finally does that. How revolutionary! I wouldn't have to paw through Thunderbird, Firefox, DevStudio, SQLServer Management Studio, a few console windows, a couple of RDP sessions, and about 5 gVim instances to find what I want to work on.
How do they resolve the names of their *own* hosts? Or, is everything hardcoded IPs?
I use a combination approach. I've a little box (FX160) that provides my core services (incl DNS). So, I can deal with "George", "Jane", "Judy", "Elroy" and "Astro" (there used to be an Orbitty but he got discarded; "Rosey", "RUDI" and "Mack" are access points) without having to remember explicit IP's.
But, I've also assinged the names with some rationale behind the choice of specific IP's. I.e., the Jetson's are x.x.x.70's and obviouosly (?) assigned in the order mentioned above. So, if the DNS host goes down, I can still access these hosts by reconstructing the IP address from the memory of just the starting address for that group of "characters".
[_The Rocky and Bullwinkle Show_ contributes "Rocky", "Bullwinkle", "Boris" and "Natasha" (names for some NAS boxes); "MrPeabody", "Sherman", and "WABAC" (my repository, of course!). Similarly, _Tennessee Tuxedo and His Tales_ contributes "Tennessee", "Chumley" and "MrWhoopee" (my "services server" -- with "3DBB" for the RDBMS!). Another "character grouping" for the laptops and portables. And, one for the X terminals. Still another for Windows hosts (The Jetson's are Sun boxen). Etc. Put a physical label on each machine "as a reminder" -- esp for the headless boxes that I can't just "ask" for their names -- and I can keep working regardless of failures or configuration errors]
I rely on Winsock for connectivity for my Compaq Portable 386! (with a parallel port network adapter!)
There's been a PowerToy for that for many years, now. I pick different wallpaper for each so I just need to remember what the background was when I was working on app. "The 4 Seasons" are a common scheme that I use: if the wallpaper had golden leaves, it was undoubtedly "Fall". Spring, Summer, Fall... so, it will be desktop #3.
I started doing this in the mid 80's on an "Opus Personal Mainframe" (add-in card that let you run SysV on a 286 host -- letting the
286 act solely as an "I/O controller"). It supported 4 "consoles" that you could access with hot-keys. Keeping track of what you were doing on each one got to be tedious. So, I'd set the foreground and background colors (ANSI 3.64 escape codes) for RED, GREEN and BLUE (RGB) with the first being WHITE! "OK, the screen was red so it must be console #2"
I use the same technique with my BSD boxes -- I've got 8 consoles defined (in addition to X sessions) so I color code the foreground and background with suitable termcap(5) entries all inheriting from the "nominal" console termcap.
Where it gets hairy is that I have multiple monitors on most workstations. And, each monitor has a "source select" switch. So, monitor #1 may be connected to output 1 of workstation 1 while monitor #2 is connected to output 2 of workstation 2 and monitor #3 is connected to output 3 of workstation 1. Workstations 1 and 2 may have different wallpaper "themes" so I may see "Summer", "Blue" (assuming the second workstation uses a color theme for workstation differentiation) and "Summer". It takes a while to sort out *what* you are looking at ("Why can't I drag this window from monitor #1 onto monitor #2???") as well as which mouse/keyboard/motion controller/tablet/etc. to use!
Particularly annoying when you're typing away and not *seeing* the text appear at the mouse cursor (of course, there are now TWO mouse cursors -- either, neither, or both may be visible and on different monitors!)... only to discover that you've been typing on the wrong keyboard and have just added a bunch of text to something completely undesired on another (potentially not visible!) monitor/workstation.
I've often wondered how confusing it would be to a casual onlooker...
I ran a FreeBSD 1.1.5.1R box for a couple of *years* in the early/mid 90's.
"Why upgrade? It's working fine!" With a UPS, there's no need to ever reboot (unlike windows that requires a reboot any time you try to change anything "significant")
I'm surprised MS hasn't scheduled a cron job to automatically initiate reboots -- rather than FIX leaks! :>
A testament to lack of memory leaks -- esp when you consider most of those processes start at IPL and just keep running (unlike windows apps that can leak and have those leaks "repaired" when the process dies and its resources are recovered!)
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.