|
I'm way too lazy to do Test Driven Dev. I like immediate satisfaction.
But I'm a fan of dogfooding. For those outside microsoft land that means using the stuff you make for your own real world things before release.
It's a royal pain though, precisely because it is good at driving bugs out of hiding.
Unfortunately, because of the complexity of Slang, any real world project with it is the total house that jack built - there's a half dozen steps to go through before reproducing anything, and sometimes bugs only crop up when i use a slang powered tool as a pre-build step which means no debugger.
I can't tell you how many times now I've had to copy the file contents of my Slang Tokenizer from github because my tool keeps breaking it. meh
Still, better now than later.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
Yes, dogfooding is both hard and valuable.
But don't overestimate it. It easily ends up as an echo chamber. "It works well for myself, then it must be perfect for the world!" I worked in one company where we hired students on short-term contracts to do both robustness and usability tests: After six months or so, they had learned to think along the same paths as the developers. They did "the right things", and reported fewer problems. We had to bring in new ones twice a year to test and evaluate the products the way users would do it, not the way developers do it.
I consider dogfooding to be the major reason for the failure of many open software projects attempting to compete against commercial end user applications: They tend to do dogfooding only. The products work fine if you are a competent software developer, but you simply can't tell an everyday secretary or dog breeder or truck driver to specify a regex for the search. Not even to use a command shell limited to 7-bit ASCII and you have to escape a lot of characters (escape characters in particular). And so on. The dogfood that pleases the developer dosen't necessarily please the customer.
One story, from the early 1980s: This office automation software, NOTIS[^], was based on an OS where users had a flat set of files - not uncommon in those days. Unix ideas were spreading, and developers were eager to offer a hierarchical file system. So they created an archive system were documents were put in folders, in drawers, in cabinets. From one of the major customers, the University of Oslo Publishing house, we heard that one of editors had found a solution to a problem: She "lost" so many documents, she spent significant resources on searching through all the folders in all the drawers in all the cabinets to find that letter she wrote yesterday. But now she had created a single cabinet, named "Cabinet", with a single drawer, named "Drawer", and a single folder, named "Folder", and moved all her documents into that folder. She no longer spent time on endless searches! All her co-workers rejoiced: That's a great idea! So within a few weeks they all had adopted the same solution, to overcome the problem of "lost" documents.
Today, it may be hard to understand that people couldn't use hierarchical systems properly, but that's how it was. While you could tell a developer to just do a "man find", the "find" syntax is completely incomprehensible to anynone without software background. Dogfooding proves that files can easily be found in a hierarchy. The developer don't see it as a problem. The customer does.
|
|
|
|
|
I agree with this, but a lot of what you said, like testers thinking like developers and so missing things is applicable to testing at large.
I don't think dogfooding is a replacement for TDD. I'm just too lazy to do TDD. 
I mean, I would if i was being paid.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
I'm also too lazy to do TDD.
And I wouldn't do it even if I was being paid!
Dogfooding is definitely the way to go. 
|
|
|
|
|
You're more honest than I am. The truth is, I probably wouldn't either.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
You just have to find which "tree" now: the one in the cloud, your PC, your "mobile", external drive, long-term storage, "junk folder", etc.
It was only in wine that he laid down no limit for himself, but he did not allow himself to be confused by it.
― Confucian Analects: Rules of Confucius about his food
|
|
|
|
|
YA THINK?
That's the second time you've now crashed and taken my VIDEO CARD with you.
That's 55 inches of NO SIGNAL fail right in my face.
Forcing me to reboot to a black screen, losing all my unsaved work.
This last time I was in the middle of a broken build.
I got locked into google back when they weren't crappy, and i used their account for everything so now I'm stuck with their browser pretty much.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
I'd look at your graphics card / drivers first, rather than Chrome - it's pretty solid for me, I can't even remember the last time it did crash.
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony
AntiTwitter: @DalekDave is now a follower!
|
|
|
|
|
My drivers are up to date. This same thing happened over two different driver revisions.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
Still probably a driver problem.
Do you have an nVidia card?
Anyway, try to disable hardware acceleration for Chrome. (In advanced setting iirc.)
|
|
|
|
|
Yeah, the first time i thought it was a fluke. This time i'm considering disabling acceleration.
I have an ancient ATI card. That might be part of the problem. I don't really care about video. The only reason I don't use an onboard device is the lack of outputs.
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
Interesting.
Some googling later. According to nVidia this is indeed a chrome problem. Solution is the same though.
|
|
|
|
|
Solution = Don't use chrome 
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
Well, blaming each other is an ancient sport.
|
|
|
|
|
Jörgen Andersson wrote: According to nVidia this is indeed a chrome problem
Without knowing any of the details, I'd still flip it around and say it's an Nvidia problem: An application should not be able to take down a driver, no matter how badly it's written (the app, that is...not the driver).
And I'm the last person ever who would try to defend Google's position.
|
|
|
|
|
Yes, but the word is "According to", and since she has an ATI-card I'm willing to give some doubt.
|
|
|
|
|
ATI...as much as I may have sworn by them in the past...never more, no matter how much they might have improved. When it comes to ATI, I'm the Linux user still frothing at the mouth because of what Microsoft may have done decades ago.
|
|
|
|
|
honey the codewitch wrote: This last time I was in the middle of a broken build.
GPU is a mysterious place. All sorts of black magick involved with DMA access.
|
|
|
|
|
To be clear, I wasn't accessing the GPU in my code. That was all chrome. My code was just some build tools I'm making
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
honey the codewitch wrote: To be clear, I wasn't accessing the GPU in my code.
Yes, we know you aren't accessing the GPU in your code. Other operating system services such as Windows Defender are now using the GPU.
Best Wishes,
-David Delaune
|
|
|
|
|
That's interesting.
Is that using as in keeping track of what's happening on the GPU, or actually using it for running stuff on it?
|
|
|
|
|
Hi,
Windows Defender utilizes the GPU for scanning and some real-time security tasks. I don't think it's fully documented anywhere but you should be able to find some third-pary commentary on the subject. It was implemented over a year ago.
Older hardware may have some overheating problems.
Best Wishes,
-David Delaune
|
|
|
|
|
Could you do a RDP connection to your machine to try to shut it down?
I think you could do it even from your cell phone...
|
|
|
|
|
If I had known in advance i could have set something up
When I was growin' up, I was the smartest kid I knew. Maybe that was just because I didn't know that many kids. All I know is now I feel the opposite.
|
|
|
|
|
If you use a RDP to shut your computer down remotely you need to use a command line instruction to do it... I still don't understand why they do this, but...
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
