AFAIK patent trolling doesn't work because of bad juries/general public but because of money and companies without tons of extra money to spare are just scared into settlements/conceding/etc.?

Who would live in East Texas?

People who don't go outside anyway :-D

Juries are not common on patent law cases. In fact there is some thought that there is no constitutional right to a jury in a patent law case.

http://patentlyo.com/patent/2014/01/jury-trials.html

Actually I partially take that back, juries may be common. But it's true that they're not necessarily a protected right.

So here's a Daily WTF. Our main website search aggregates stuff from all over the company. I just found out that a lot of it is driven by scraping our own internal and external sites, which is then sent through multiple processing/parsing steps, then eventually dumped into a Lucene index.

The reason for the site scraping is it's easier than getting those other groups to cooperate by setting up a data feed. o_O

Yeah that's some pretty heinous bull****.

Yeah I was like - please don't tell me we're site scraping because it's easier than setting up a data feed. Yep.

I've seen a lot of the reverse though too. Like finding out that some other project has been connecting directly to your database and running queries (instead of using your API or requesting one). The problem usually is: the data is not necessarily what it looks like, plus, if I don't know you're doing it, I will **** you over by changing stuff.

Oh and at my last job a big part of my job was writing these functions to calculate a "score" for something. What the score was isn't really important but each customer had a different scoring routine and we made custom ones for each of them, 3-10 a month.

A division of the company needed some marketing data, so they had their data guy re-produce the scoring routine (independently) from the data they had access to (which was... not of the highest quality to begin with). It would have been *really* trivial for us to just dump the data to them nightly or weekly or whatever, instead of burning an entire person to work on it non-stop to produce wrong results.

In fact right before I left I added a url they could go to, to dump the data. It took me about an hour. The guy who used to be in charge of generating was like "we could have had this all along?" Yep.

i've dealt with very similar things myself, and it's funny because i don't even view these as particularly wtf anymore (tho ofc they are absurd in a way), because they are explained so well by politics and human behavior. basically, all just instances of conway's law.

"Politics" is a dirty word at my company. I generally use "decisions made by people above my pay grade" as a replacement.

Eh, mostly it's people at my pay grade.

And I guarantee that I've done some stuff that people find equally frustrating, because I had a god damn job to do. Not long before I left UA I needed to make one of those fancy double email optin things for euro customers. We had a "service" to do double optin stuff. However, whoever made it did not make it configurable *at all* and it was not going to work for me. For example it had a single setting per user - but I needed to support dozens of seperate optins per customer. In fact I even needed to support multiple settings per campaign.

I could have
a) come up with some kludge to make it work, like, pretending that "user_id" was really "user_id+campaign+customer" or something equally ****ty
b) fixed it myself - hello 2 weeks of bull****, no thanks
c) convinced someone else to fix it - probably not gonna happen, whoever made it just threw it together and released it into the wild and went on to other ****.

So I made my own.

Code-reuse is a myth anyway, mostly

certainly what you are talking about, but i mean it even more broadly.

eg, it could be that the engineer who made the scraper doesn't like the one who could provide him a nice API because he flirted with his gf once at an office party, so instead of asking him he just makes the scraper.

or even more innocently: the guy who can make the API in 30 minutes happens to be on vacation the rest of the week, but i need to finish this today, so **** it, i'll spend 3 hours making a scraper.

etc etc.

so much this.

i wouldn't say that it's a total myth, i'd just say that it's a) really hard to break things apart the way they should be and b) often requires more upfront work to do so, even if you know how.

once you realize that it requires both smarts and lack of laziness, it becomes clear why it's rare, and why just making it over is so often easier than attempting to reuse a 70% right solution.

My daily WTF is totally What.The.****.

There's three people involved in this story, call one JJ, the other MM, and the other is me.

JJ gets a free computer the other day. It is old as dirt, running on Pentium M, and about 2x the size of my own computer. It is safe to assume this doesn't have stellar specs, but hey, it is a free computer.

It has Ubuntu on it, but he forgot all of the passwords for the computer. Uh... I suggest handing this off to MM because he knows a thing or two about running Ubuntu server and I'm fairly clueless on this. I suggest removing the current whatever and reinstalling, but defer to MM.

MM takes a look at it and thinks the same thing, but he and I decide to have a bit of fun with it first. How do you reset a lost password for root on this when you can't login as any user and only can be a guest?

Turn on the computer, when grub comes on, hold down Shift then press "E" to edit the OS you want to edit. He and I are skeptical this could be real, but why the **** not? I do as required and what do I see?

oh.... then I have to enter:

this can't be real... enter the following:

it worked. Please tell me the server version is more locked down than this!

I go the OS and it is is Ubuntu 14, complete with Unity, slow as molasses in winter, and yes, it is even running the Amazon Spyware. Not that the latter mattered since we weren't on the internet.

I explore the computer, and there are about a dozen oddly named users, none found in /home. I explore some more and it doesn't make any sense. The computer appears to be empty, but "du -h" shows over 5g of something on it. Instead of upgrading two versions, removing the Amazon Spyware and trudging around searching for other bloatware, MM burns an ISO and we start loading Lubuntu on the computer.

Lubuntu does not come with a Wifi connection by default. We have to enter some values in some strange place and reboot. We do so and... no Wifi, no WiFi connections manager, no nothing.

Why? According to the Lubuntu docs, there is no WiFi by default because the OS is for "old computers, and old computers don't have Wifi." Uh... this one is probably 10 years old and has WiFi. WTF?

In order to turn on WiFi, you have to go into the command line and enter archaic commands (I have no idea if this is true considering the original plan didn't work), but the password created during installation doesn't work on the command line. We have to figure out how to get into root, but apparently some password is default and it isn't the empty string.

MM searches online reads the real gold for me: "don't use sudo. Sudo is bad."

What?

"That's what the official documentation says."

What?

MM probably thought about what more ****ed up **** we were going to ram into and decided to call it quits.

(I've never had a positive experience with Ubuntu. This didn't improve my opinion of it at all)

I've been using linux since it came out, more or less. You've always been able to break into it very easily, as long as you have control over the boot sequence. There is literally nothing you can do about that aside from maybe whole disk encryption. I was the network admin at our college while I was a student there and I used to demonstrate to anyone who cared that all you needed was a floppy disk with slackware on it to break into any linux computer that had a floppy drive.

Most of the major distributions these days are fine. I have unbuntu on a few machines in the house, the rest are debian. I have never used "lubuntu"

Wifi on linux has always been something of a crapshoot. This appears to be due to a lack of standards in wifi hardware combined with wifi hardware manufacturers not being willing to give enough info to make good wifi drivers. I remember than in the mid 2000s the best way to get a working wifi driver was actually to use the *windows driver* along with some magic software that let you use the windows driver with linux. Even then, you couldn't do wireless-ethernet bridging, I don't remember why.

Lubuntu is Ubuntu with an LXDE desktop instead of Unity, and apparently no easy way to set up WiFi. The WiFi on Ubuntu regular showed up and connected. Just that Lubuntu has no WiFi by default. Idk.

I'm aware there are ways to take over a system with a disk, but this way seems a bit too easy.

Can't you password protect the bios and grub to prevent this?

Uploaded resume to dice without phone number because why not, 24 hours later https://www.youtube.com/watch?v=yDbvVFffWV4

I hope you have better luck than I did having actual stuff to do. I have 2 jobs and still nothing to do but vague, amorphous, open-ended, non-technical tasks that for the most part no one cares if I do or not. But the side job might be getting close to actual stuff.

I have stuff to do, I just don't want to do it and its a complete PITA and not getting me anywhere and I don't have any agency. So we'll see what's out there why not.

Had to reinstall my OS today. Not exactly sure what happened.

Basically, there was a long NCQ queue, which means something was trying to get written to disk and it failed to write. After trying to reset and write all ones, it still didn't work.

Either it is a totally freak accident (and buggy LXQT), a hard drive wire is burning out and not writing correctly, or, the hard drive is taking a dive.

Didn't lose much that I cared about, but still incredibly irritating. Trying to install everything is a whole other story.

What does the SMART data for the drive say?

Hopefully I'm not totally screwing this one up:

Does PASSED mean anything here?

I'm wondering about the line: "Device is: Not in smartctl database [for details use: -P showall]"

I am using a PNY SSD, 480G

Passed means, hdd passed smart test which is good but in general doesn't tell you much. Db line is nothing important. If you start getting write/ read errors in system logs thats not good.
If you got time i d look into downoading a live distro (i like knoppix for variety of hw support) boot of it and perform "secure erase" on the ssd. Make sure its the right drive Then reinstall everything thats not that important and monitor logs for few days. Or if data is imporfant, just replace the drive.
Here is a link on ssd "secure erase" https://wiki.archlinux.org/index.php..._cell_clearing