Coding Horrors
-
I suppose most of us have seen this article:
There's a comment on there:
Jeff Atwood has a website called "Coding Horrors" filled with examples of what happens when people with no skill try to code. If this moves forward there will be even more grist for his website.
Did he mean discourse.org or thedailywtf.com?
-
From Jeff's article.
You had to learn the right commands to get them to do anything at all. In other words, you were forced to become a computer programmer in order to be a
computerDiscourse user.FTFJ.
-
thedailydiscourse.com
is available...Just saying...
-
It seems that the only people associated with Discourse who aren't forced to become skilled programmers are Discourse Developers. Now I see that Attwood's doing this on purpose.
-
I was going to defend them somehow but then I reminded myself about 3389dae361af79b04c9c8e7057f60cc6.
-
.... as usual, he has a point (programming without reading/reasoning is worthless) but his delivery is so disjointed and full of wibbly wobbly bloggy woggy stuff that his point is sufficiently obscured as to be worthless as the media won't read and understand his point and sure as belgium the public won't read it, let alone understand it.
-
I don't understand, why does the post source have more characters the the actual post? Two asterisks and
3389d
just vanish?
-
I don't understand
Discourse.
why does the post source have more characters the the actual post?
Discourse.
Two asterisks and 3389d just vanish?
Discourse.
-
thedailydiscourse.com
so is "thedailyintercourse.com" aswell as "thedailybumraping.com" Guess which one I'm buying when my next pay cheque comes through.I think he has a point about not wanting more people in this industry though. There's already enough egotistical fuck ups in it. He been king shit of turd mountain.
-
IMO the bigger issue is that so many companies are incapable of filtering out the bad programmers. About a fifth of TDWTF is just stories about people who should never have been hired.
-
About a fifth of TDWTF is just stories about people who should never have been hired.
I'd say that us talking about Discourse takes up more than a fifth.
-
Discourse : building communities by creating an enemy that everyone hates.
The system works.
-
I'd say that us talking about Discourse takes up more than a fifth.
"TDWTF" != "WTDWTF"
-
-
Also, I really love that it does that. There's still some funky shit happening with the parser, they've just band-aid-ed it. Sorta... Kinda?
-
Depends. If you go with regular spirits, sure. But spiritus can get you through on a split.
-
The fact that they have to patch things around with MD5 should have hinted to them that parsing this whole soup with regexes might not be a great idea.
But oh well.
-
This goes all the way down, doesn't it? The more I find out, the more I am frankly amazed that this thing even close to works.
-
Learning to talk to the computer is the easiest part. Computers, for better or worse, do exactly what you tell them to do, every time, in exactly the same way. The people — well . . . you’ll spend the rest of your life figuring that out. And from my perspective, the sooner you start, the better.
He really should take his own advice and learn how to talk to people.
-
kinda half-hoping we're staying with Discourse for three more months, just so I can reuse a joke from that thread... (not Discopedia, the other one)
-
Also, I really love that it does that. There's still some funky shit happening with the parser, they've just band-aid-ed it. Sorta... Kinda?
Bug report: * causes 3389dae361af79b04c9c8e7057f60cc6
Reason: * is being MD5 hashed into 3389dae361af79b04c9c8e7057f60cc6
Expectation: * shouldn't be 3389dae361af79b04c9c8e7057f60cc6Discourse staff replies
We have fixed your unexpected text. We now take , concat ~ to it, and do MD5("~"). You should no longer see 3389dae361af79b04c9c8e7057f60cc6.
-
Markdown: Where the unexpected... is expected
-
I thought something was <wrong>about that...
-
ae361af79b04c9c8e7057f60cc6**.
zomg how long ago was that bug reported? It must be at least 6 months?
-
I thought their reply was more along the lines of " developer, why are you trying to break the software? Real Users never input MD5. "
-
Jeff Atwood has a website called "Coding Horrors" filled with examples of what happens when people with no skill try to code.
No.
-
It's true though; his blog uses discourse for comments, so you have a minimum of one example per blog post.
-
Fucking wow.
-
zomg how long ago was that bug reported? It must be at least 6 months?
2 months and 2 weeks, actually.
-
a fifth of TDWTF
I realise this is a separate statistic, but I wonder how much of the actual stories1 are not about people who should not have been hired.
1. I mean with actual WTFs. Don't try skewing the data with stories like "I made a cake." I'm watching you.
-
Well, the way I was estimating was stories where at the start of the story, the person who committed the WTFery was not yet working for the company. Obviously if you are less specific, then over half of the articles are about people who should never have been hired.
-
.... as usual, he has a point (programming without reading/reasoning is worthless) but his delivery is so disjointed and full of wibbly wobbly bloggy woggy stuff that his point is sufficiently obscured as to be worthless as the media won't read and understand his point and sure as belgium the public won't read it, let alone understand it.
It's not helped by the fact that like a lot of people these days, he equates "programming" and "coding" and then complains that the resulting mess doesn't contain the rest of what's actually included in "programming". All the stuff like reading, reasoning, design, and so on, that aren't part of actually banging on the keyboard in a code editor are, in his weird world, not part of programming either.Sigh.
-
So… you're saying that Jeff is a code monkey with a big mouth and not a software engineer? I could buy that.
-
@Gaska said:
ae361af79b04c9c8e7057f60cc6**.
zomg how long ago was that bug reported? It must be at least 6 months?
Mmmm...yes, at least. Known for over 10 months to discodevs:
https://what.thedailywtf.com/t/literally-wtf/4984/10?u=boomzilla
Reported that same day at fail.d:
https://meta.discourse.org/t/cannot-escape-in-bolditalic-markdown/22280?u=boomzilla
-
I'm sorry my mind refuses to understand what 3389dae361af79b04c9c8e7057f60cc6 does. It seems my internal safeguards against insanity have been activated and this cannot get past. Maybe someone could rephrase it in a way the safeguards cannot detect and let me escape this prison of sanity
-
md5(*) == 3389dae361af79b04c9c8e7057f60cc6
-
Relatedly:
md5(_) == b14a7b8059d9c055954c92674ce60032
-
b14a7b8059d9c055954c92674ce60032
-
The real question is why anything in the body of our posts needs to be hashed or compared to hashes or anywhere near the concept of hashing.
-
It was a way of escaping ("hoisting") asterisks out of content without having to write some obscenity like
\*\*\*(.*?)(?<!\\)\*\*\*
(which... can you even do that in both Ruby and JS regex?)
-
It was a way of escaping ("hoisting") asterisks out of content without having to write some obscenity
I prefer to use some of the more exotic parts of Unicode for that sort of thing, on the grounds that they're much less likely to be discovered by accident, and they're a single character (though multiple bytes ) and so are easier to handle later on.
Yes, I've done some horrible RE stuff in my time.
-
Let's watch as we approach "3389dae361af79b04c9c8e7057f60cc6****"
***3389dae361af79b04c9c8e70
***3389dae361af79b04c9c8e705
***3389dae361af79b04c9c8e7057
***3389dae361af79b04c9c8e7057f
***3389dae361af79b04c9c8e7057f6
***3389dae361af79b04c9c8e7057f60
**3389dae361af79b04c9c8e7057f60c
3389dae361af79b04c9c8e7057f60cc
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6
3389dae361af79b04c9c8e7057f60cc6It's a disco ntinuity
BUT WHY * TELL ME WHY!
-
The original complaint was that
***\****
would produce 3389dae361af79b04c9c8e7057f60cc6, so they fixed that problem the only way they knew.
-
Thanks! I'll just let my inner child do what comes naturally
It was a way of escaping ("hoisting") asterisks out of content
WHY?The original complaint was that
***\****
would produce 3389dae361af79b04c9c8e7057f60cc6WHY?
-
New discodev kit includes a pair of complicator's gloves.
-
Open source people don't think there's any difference.
-
WHY?
replace('\*', md5('*')); //get this out of the way, so we don't screw it up later replace('**[...]**', '<b>[...]</b>'); replace('*[...]*', '<i>[...]</i>'); //put back what we... oooh, kitty! lemme go play with the kitty, brb
-
s/[...]/.*/g
-
WHY?
It makes sense when you remember that the discodevs probably don't know how to even write a typical parser, so they just throw regular expressions and shitty hacks at the problem until it kinda sorta resembles a parser.
It's pretty funny though, as jeff himself has previously said regular expressions are no substitute for a real parser.
-
I prefer to use some of the more exotic parts of Unicode for that sort of thing, on the grounds that they're much less likely to be discovered by accident, and they're a single character (though multiple bytes ) and so are easier to handle later on.
Yes, I've done some horrible RE stuff in my time.
I've done it by using ASCII control codes, after simply stripping all user-entered control codes apart from tab and newline. They're single character and single byte which makes processing them with regexes dead easy.
@accalia has teh codez if you want to look under that rock. Google is not helping me find the Dicksauce copies.