|
|
I was not active on CP for quite some while. Now that I looked into some of my articles/tips, I noticed that some a bit "advanced" formatting got completely screwed up in the last two years or so.
E.g. highlighting, line numbering, background/foreground color got corrupt.
See Invent your own Dynamic LINQ parser[^] at the end of the first pre-block (span...highlight is verbatim instead of html/css directive).
I also have on several places used the foreground/background colors to emulate console output with black background and white text. See the same article above in the pre-block after "The output is:".
The feature of line numbering of code snippets also got silently removed. Why? I some times refer in the text to these lines.
Basically all my articles/tips are affected since I took benefit of these features those days when they were functional...
Any remedy in sight?
Thanks
Andi
|
|
|
|
|
Try as I might, I can't seem to get highlighting back into that pre block. It used to be there though, correct?
Thanks,
Sean Ewington
CodeProject
|
|
|
|
|
Indeed the highlight (yellowish background) seems to cause this.
The same defect is seen in version 2+ code blocks where I've tried to highlight the changes compared to previous versions: Properly executing database operations
|
|
|
|
|
Found and fixed and will deploy once I've finished testing
cheers
Chris Maunder
|
|
|
|
|
Thanks!
Looking forward!
Cheers
Andi
|
|
|
|
|
|
Hello Chris,
the highlighting works at places where the article sources did not get screwed up themselves somehow. E.g. see Invent your own Dynamic LINQ parser[^] at the first pre-block. The source view shows that the embedded highlighting in the pre-block got converted over time into plain code... (likewise other highlighting in the same article and others of mine).
The other similar issue is about background color. I did use to simulate console output by giving the foreground color white and the background color black in some pre-block. This used to work in the (distant?) past. Now, it looks like it's kind of filtered: the foreground color is still respected, the background got removed completely. You seem to have some stripping/processing/whatever machinery that processes ***existing*** articles once in a while (my interpretation of the situation). E.g. see Escaping in C#: characters, strings, string formats, keywords, identifiers[^], the pre-block after "results in" used to be started like <pre lang="text" style="background:black;color:white;">... but now is truncated to <pre lang="text" style="color: white">.
Finally, the pre-block attributes countlines and countstart got stripped off somehow from the sources... E.g. see my History section, V1.2 comment in Invent your own Dynamic LINQ parser[^].
Are these any deprecated features?
Regards
Andi
|
|
|
|
|
Looks like lots of your &span> tags were being HTML encoded. I've fixed these.
Background colour should be set using background-color, not background. We don't allow background because it's too difficult to sanitise safely.
With regards to countlines: these were being stripped out but I've added them back to the "allow" list. Which blocks do you want to have with line counts?
cheers
Chris Maunder
|
|
|
|
|
Often many comments are left on a question in Q&A, many of them asking the same thing, clarification or commenting the question.
On the other hand, the OP may not know the system and how the author of the comment is notified (or is not notified) so the may just leave a new comment, not a reply.
So when the OP comments his own question (which probably is an answer to previous comments), why wouldn't all the authors of the comments be notified at the same time?
Optionally there could be a check box to select "Notify all commentators" and this would be selected by default if the comment isn't a reply....
|
|
|
|
|
About the same problem but a different suggestion.
When replying to your own forum message there will be a notification:
"You are about to reply to your own message. Is this really what you want?"
How about something similar with comments on questions, answers, and comments?
Probably with additional text about how to reply to other comments and edit the question.
|
|
|
|
|
I was thinking about this but what if the question has 5 similar comments, all requesting fore more info. Wouldn't it be nice to inform all at once.
Often nobody gets informed because the reply comment is written on the 'top level' or only one person is informed; the one for whom the reply is written to.
|
|
|
|
|
That would be an - nice to have - additional option.
But I think that is rather seldom compared to posting comments at the wrong place.
If I wrote such comments asking for clarification, I usually visit the question again. If the question is improved, it is brought back to the top of the list so that I usually recognise it without getting informed by mail. So giving a hint for improving the question makes the most sense.
I wasn't going to deny your suggestion. I just thought there were other options that might be helpful too.
|
|
|
|
|
Jochen Arndt wrote: But I think that is rather seldom compared to posting comments at the wrong place.
That is very true.
Jochen Arndt wrote: If the question is improved, it is brought back to the top of the list so that I usually recognise it without getting informed by mail.
That is also true, unfortunately often the enhancements are not made in the question but posted in the comment (a mile long comment with all the code)  Perhaps comments on the questions should also bring the question back to the top of the list??? At least if they are from OP
Jochen Arndt wrote: I wasn't going to deny your suggestion. I just thought there were other options that might be helpful too.
I didn't take your comments that way, on the contrary, it's good to have multiple opinions 
|
|
|
|
|
A message like "You are posting a new comment to your own question instead of replying to an existing comment. Are you sure this is what you want?"
There are two kinds of people in the world: those who can extrapolate from incomplete data.
There are only 10 types of people in the world, those who understand binary and those who don't.
|
|
|
|
|
Done
cheers
Chris Maunder
|
|
|
|
|
That's an excellent suggestion, and one which we're planning on adding to the forums, too. You post a message in a thread and you get notified whenever something else in the thread changes (available opt-out, of course)
cheers
Chris Maunder
|
|
|
|
|
I see a lot of situations where a newly posted question is downvoted immediately. Sometimes the reason is obvious, sometimes not. But very seldom any message is posted explaining why the question is downvoted.
So what every now and then happens is that a moment later the OP adds more details to the question and it becomes quite good one, but the downvoter is not likely to return and re-vote...
I see two downsides on this:
- In a situation where OP really tries to cooperate, he gets 'rejected' and probably confused
- When the question is enhanced afterwards, the downvote gives wrong signals to later readers.
I understand that the downvotes are good and necessary to distinguish poor questions but the outcome is probably not what's expected in the long run...
Perhaps a time delay? Voting begins after 30 minutes or after first comment???
|
|
|
|
|
Mika Wendelius wrote: the downvote gives wrong signals to later readers.
I rarely notice downvotes since they don't mean anything anyway. However, sometimes I see one downvoted that I think does not deserve it and then I upvote it.
There are two kinds of people in the world: those who can extrapolate from incomplete data.
There are only 10 types of people in the world, those who understand binary and those who don't.
|
|
|
|
|
RyanDev wrote: However, sometimes I see one downvoted that I think does not deserve it and then I upvote it.
I try to do the same if the question good.
Regulars at CP understand the voting system and how reliable the votes are but I'm mainly thinking of people surfing, looking for an answer to a problem. A downvoted question may drive them away even if it's good and has good answers...
|
|
|
|
|
Mika Wendelius wrote: A downvoted question may drive them away even if it's good and has good answers
It might. For me, when I google or search I look at the upvotes on answers, not downvotes on questions. But I am unique. 
There are two kinds of people in the world: those who can extrapolate from incomplete data.
There are only 10 types of people in the world, those who understand binary and those who don't.
|
|
|
|
|
"Surfing"?
That's a ridiculous notion; someone looking for an answer getting a codeproject hit and not being a member would have to sign up to become a member in order to vote.
|
|
|
|
|
RedDk wrote: "Surfing"? That's a ridiculous notion
I probably picked a wrong term, most likely browsing would be better.
RedDk wrote: someone looking for an answer getting a codeproject hit and not being a member would have to sign up to become a member in order to vote.
That's very true but reading doesn't require membership. My point is that, regardless of being a member or not, low votes on good questions may unnecessarily drive both questioners and readers somewhere else.
I admit that poor questions should be marked but the question is, should a question in Q&A be given some time to improve before being down-voted.
|
|
|
|
|
Interesting ...
I DO ignore low ratings ... but the most ignorable trait of most posts that are low caliber is the typical poorly worded or illiterate think-speak which is going to get a zero vote from me everytime without me looking for me feet.
I see the main problem in QA as being the engagement of an illiterate post in any way shape or form, comment or otherwise (NEVER an answer).
Hence my "Remove this post" penchant that only the hamsters know about!
|
|
|
|
|
There's an overarching issue here.
If we have poor questions then people can't be bothered to hang around and answer them. The better the question, the easier the poster has made it to answer, the more likely it will be answered.
Instead of tweaking voting, I'd prefer to focus on getting the questions framed correctly in the first place. To this end it may even make sense that if a high rep member downvotes a question (and maybe they are the only ones that *can* downvote) then the question goes offline, the poster is notified, and standard tips on improving the question are included. The poster either responds and fixes, or ignores it and the question remains out of harm's way.
cheers
Chris Maunder
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
