Archive for the 'Tech-y' Category
Recently, I encountered my quarterly issue with using an iphone with outlook. I say “quarterly” because this combination is rickety and unreliable, and breaks several times a year, which is to be expected when using software form the two largest companies on the planet. What are all those thousands of brilliant people on posh campuses doing with their time, exactly, if they can’t keep this stuff working? It boggles my mind.
That ranting intro is my way of saying that if you’ve arrived here trying to solve your iPhone and outlook problem, it might be similar to the issue described in this post, or it might not. Good luck. This happens to me multiple times a year, and the problem is different every time. One thing i can be sure of: the problem is never my fault; solving these issues is my burden for being foolish enough to stop using pen and paper.
Issue: Some event and contact info does not synchronize between the phone and outlook. I can’t see any consistency in how or why some makes it and some doesn’t. However, what I see on icloud.com does match what’s on my phone; it’s just that what I see on outlook doesn’t match what’s in icloud.
Solution: in the windows icloud application, sign out. Let it erase all of the icloud information on the local windows computer (after verifying that it is, indeed, in icloud). Sign in and let it download. Now, all will be right.No comments
In the case of logging in as a normal user and invoking “su – root” I found that Mac OS 10.8.5′s bash was ignoring .profile and .bash_profile; I was unable to change root’s $PATH by editing those files. What did work was editing /etc/paths. After exiting the root shell and entering again with “su – root” the new path was present.No comments
At work I routinely have to ssh from host A to host B and then to host C. It is not possible to establish a direct link from A to C, so I’ve been manually establishing a connection from one to the other, which is annoying. Today I finally got around to setting up a better solution, which depends on having a unix-like system on all nodes. Host A is running Windows with cygwin, and hosts B and C are running Mac OS and Linux. Another assumption is that you have the same user name on all nodes, although there are ways to get around it if you don’t.
1) Add these lines to ~/.ssh/config (which you may have to create).
Host hostC.domain.edu hostC
ProxyCommand ssh hostB -W %h:%p
Now, when I type “ssh hostC” on my PC, it hops through hostB and logs into hostC automatically with no further typing.
What is happening is that when you ssh to hostC, ssh substitutes another command for /bin/sh on hostB (which is normally executed by default), and forwards stdin and stdout to this new command (that’s what the –W is for), which is an ssh to rsndds. The effect is to hop through hostB. Because I’ve set up keys without pass phrases on all machines, no password is required. If you don’t have keys set up, it will still work, but will ask for passwords.
If you want to use pass phrases, you can use ssh-agent on hostA and hostB, which will ask only once and then store the keys in memory until the next reset of ssh-agent (probably a reboot).
If you have a different user name on hostB, simply specify it like this:
Host hostC.domain.edu hostC
ProxyCommand ssh user@hostB -W %h:%p
You can create as many of these special entries in your config file as you wish, each specifying special rules for creating connections to your unique networks. Isn’t ssh cool?
When I started figuring out how to do this, I thought of it as “tunneling,” but technically that’s probably not correct; that name is given to using ssh for encrypted port forwarding. Nevertheless, you can think of it as a tunnel, allowing you to ssh from one machine to another using an intermediate machine, all without your intervention.No comments
I have an Asus RT-N16 router and after years of perfect service encountered intermittent operation; the unit would operate for hours at a time and then turn off randomly. It would turn itself back on after some time, perhaps 20 minutes. This behavior continued to deteriorate until it would only stay on for perhaps 10 seconds. It felt like an overheating problem, but I was incorrect. As with all problem-solving nowdays, I first consulted the ‘net and found this link.
The capacitor pictured in the link looked perfect, and when I used my ‘scope to examine the quality of the DC power rail it looked fine, but having lots of spare capacitors and nothing to lose I replaced it anyway, and it was fixed!
Moral of the story: capacitors don’t have to be visibly damaged to be faulty. They don’t even have to stop working to need replacement! Also, the quality of the DC power rail is a tricky thing to use, because if the power supply is adequate for most purposes but fails under a particular kind of stress, you will have to a) figure that out and b) observe it while that even is occurring; this can be an elusive thing. It’s possible that had I measured its capacitance, it would have been fine; it could just be that its ESR had risen to unacceptable levels and failed under the high-current demand imposed when the unit’s transmitter quickly used a lot of power. This is a typical failure mode for capacitors; as they get older they still function, but have poor ripple rejection despite having their original value in Farads.
I didn’t have a capacitor of the same physical size, so I removed the defective one, soldered jumper wires and installed a physically large 2500 uF unit where there was room in the chassis a few inches away. It’s been working perfectly for weeks now.2 comments