Bugs and Suggestions

Richard Deeming27-Jun-17 4:05

Richard Deeming

27-Jun-17 4:05

Chris Maunder wrote:
a button at the bottom of the page "mark all read"

Would that apply to the entire site, the current forum, the current page of the current forum, or something else?

Whichever you choose, I think it's likely to cause confusion. Smile | :)

How about an option to temporarily override the "last visit" date/time? And store it in a session cookie, so that restarting the browser restores the default behaviour - no need for an "unlock" button.

"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer

Chris Maunder27-Jun-17 4:33

27-Jun-17 4:33

Richard Deeming wrote:
Whichever you choose, I think it's likely to cause confusion.

Me too

Richard Deeming wrote:
How about an option to temporarily override the "last visit" date/time?

I'm guessing this means "when I open a new page, I want a button that would set the "last visit" value back to the value it was before I visited". This is possible.

But...

Richard Deeming wrote:
And store it in a session cookie

This is the kicker. The original message discussed an issue where the browser was left open for long periods. After 20 mins your session expires so if a browser was left open then the session expires and nothing's stored. We could store the value in local storage or a JavaScript var and carry that around, but that seems...clunky.

What would be *really* good is a per-message flag that says whether or not the message has been read - which is hugely resource intensive and not worth the pain and cost.

cheers
Chris Maunder

Richard Deeming27-Jun-17 5:13

Richard Deeming

27-Jun-17 5:13

Chris Maunder wrote:
I'm guessing this means "when I open a new page, I want a button that would set the "last visit" value back to the value it was before I visited".

I was thinking more of a pop-up to enter a custom date and time for the "last visit" value. Possibly triggered from the "indicates new messages since" area.

But obviously I appreciate these things are much easier to say than they are to implement! Big Grin | :-D

Chris Maunder wrote:
After 20 mins your session expires

No, that the server's session. I mean a "session cookie" (aka "in-memory", "transient" or "temporary" cookie) - one with no expiration date set. They're stored until the browser closes, and aren't connected to the server's session. Smile | :)

Chris Maunder wrote:
What would be *really* good is a per-message flag that says whether or not the message has been read - which is hugely resource intensive and not worth the pain and cost.

Agreed.

"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer

Chris Maunder27-Jun-17 8:49

27-Jun-17 8:49

Richard Deeming wrote:
a pop-up to enter a custom date and time for the "last visit" value. Possibly triggered from the "indicates new messages since" area

That might work. And yeah - a session cookie (as opposed to server session storage) would do it.

I just wonder how many would actually use this feature...

cheers
Chris Maunder

Richard Deeming27-Jun-17 9:23

Richard Deeming

27-Jun-17 9:23

Chris Maunder wrote:
I just wonder how many would actually use this feature...

If only there was some way of posting a poll on the homepage to ask! Big Grin | :-D

"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer

Chris Maunder27-Jun-17 9:35

27-Jun-17 9:35

Now why didn't I think of that. What an awesome feature that'd be.

cheers
Chris Maunder

More post information shown at "top-level" on QA ?

Peter_in_278027-Jun-17 14:18

Peter_in_2780

27-Jun-17 14:18

I was envisaging a small button near where it says "new messages since..." Label? something like "update" or "set to now". And a profile page checkbox to opt in to the whole thing. That seemed to me to be minimal impact, given that you obviously keep a per-member timestamp anyway. I just got annoyed by the auto-update mechanism (whatever it is) marking a whole bunch of stuff "not new" when I hadn't browsed the fora.
Looking at the other discussion on this thread, I never envisaged you'd consider per-member per-message. O(n^2) and all that.

Cheers,
Peter

Software rusts. Simon Stephenson, ca 1994. So does this signature. me, 2012

BillWoodruff24-Oct-15 8:58

BillWoodruff

24-Oct-15 8:58

I would find it useful if the member-name of the poster were shown after the information about the last revision/update time.

imho, this would be useful for detecting a series of posts by the same OP, etc.

thanks, Bill

«I want to stay as close to the edge as I can without going over. Out on the edge you see all kinds of things you can't see from the center» Kurt Vonnegut.

Re: More post information shown at "top-level" on QA ?

Chris Maunder24-Oct-15 10:09

Re: More post information shown at "top-level" on QA ?

24-Oct-15 10:09

There was actually a deliberate decision not to do this because we want questions answered (or voted or reported or even closed) based on the question, not the author.

To view all questions by a single author use http://www.codeproject.com/script/Answers/MemberPosts.aspx?tab=questions&mid=">http://www.codeproject.com/script/Answers/MemberPosts.aspx?tab=questions&mid=[member id]. However, I'm guessing you're looking at spotting, within the QA list itself, a series of questions from the same member?

cheers
Chris Maunder

BillWoodruff24-Oct-15 20:01

BillWoodruff

24-Oct-15 20:01

Chris Maunder wrote:
However, I'm guessing you're looking at spotting, within the QA list itself, a series of questions from the same member?

Yes.

I note that on the C# language forum, you do always see the author's name (of course, that forum is in a different format than QA forums). The difference in the way QA forums are structured compared to language forums, like C#, is the source of my hypothesis that the original intent here was to have the language forums serve/perform a different purpose than the QA forums, within the context of the whole site.

That's my excuse for sometimes decrying the slip-sliding over time of the C# forum into a second QA forum, rather than a place where language issues get discussed in depth. However, there are threads in the C# forum where discussion does go marvelous deep into language issues, and, certainly, many threads there have a 'technical depth' you see more rarely in QA (in part, thanks to the zeal of our current crop of QA's reputation-uber-alles hotshots ?).

On the C# language forum, I 'enjoy' the fact that it is easy for me to spot and ignore the on-going chain of messages from the CP member who's been on the site over 13 years, and never done anything but ask questions, and who, to my knowledge, has never thanked anyone who replied to his messages, or voted-up anyone who replied, and whose questions suggest that he has CP doing the work of the business he claims to own ... for him.

But, please don't think my enumeration of quibbles here is anything more than the finicky nit-picking of someone who, admitted to heaven, after a while started noticing a few pimples on 'angels' Smile | :)

thanks, Bill

«I want to stay as close to the edge as I can without going over. Out on the edge you see all kinds of things you can't see from the center» Kurt Vonnegut.

Re: More post information shown at "top-level" on QA ?

ZurdoDev26-Oct-15 4:00

ZurdoDev

26-Oct-15 4:00

Chris Maunder wrote:
There was actually a deliberate decision not to do this because we want questions answered (or voted or reported or even closed) based on the question, not the author.

That is a good point. I also agree with Bill, it would make it easier to deal with questions.

There are only 10 types of people in the world, those who understand binary and those who don't.

Suggestion: Featured Articles

newton.saber24-Oct-15 7:10

24-Oct-15 7:10

How about adding an algorithm that displays featured articles which checks to see if the author has logged in within the past X days.

That way, authors who haven't visited CP in X days don't get their articles displayed.

Also, maybe a date limit on those which only goes back 2 years at most. Maybe only 1 year.

Does anyone read these suggestions?

Jörgen Andersson24-Oct-15 8:21

Jörgen Andersson

24-Oct-15 8:21

And may I ask why?
Does the fact that the author isn't active make an article less worth?

Wrong is evil and must be defeated. - Jeff Ello

newton.saber24-Oct-15 8:28

24-Oct-15 8:28

I guess because I think of the articles as the beginning of communication on a topic.

I think this site is dedicated to instruction and discussion so if the original author hasn't logged in within some large number of days it means it will probably be difficult to discuss her intentions with the article.

Although I understand there could be good articles with good content that could be missed. But for those types of things it is probably better that someone is searching specifically for them anyways.

Thanks for discussing this. Thumbs Up | :thumbsup:

I hope to hear your insight on this.

Jörgen Andersson24-Oct-15 8:51

Jörgen Andersson

24-Oct-15 8:51

Yes, I get your point, but some articles have such a limited subject or are so well written that there's not much to discuss.

That said, please don't misunderstand me the wrong way, I'm quite disappointed in the lack of response on my own articles.

Wrong is evil and must be defeated. - Jeff Ello

newton.saber24-Oct-15 9:14

24-Oct-15 9:14

I think you are backing up my point though. You would get more attention on your articles because they would bubble to the top more often, because you log on often. There could be a whole weighting thing with author login, article vote rating, number of views.

It seems like that would better define "Featured Articles".

Edit:
I will also make my way to your articles and check them out.

BillWoodruff24-Oct-15 8:52

BillWoodruff

24-Oct-15 8:52

I think that some articles (very highly rated by many voters, history of revision by the author over time, history of good support by the author over time) are CodeProject's "Jewels in the Crown," and deserve to get featured in rotation, or whatever, no matter how many years mileage they have.

If an article has not been highly voted by many voters, and is more than some months old, I don't think it should be featured.

Those are my preferences, but, as I write this, I know I don't know what method is used to select featured articles now; so, perhaps we should ask CodeProject staff about that ?

cheers, Bill

«I want to stay as close to the edge as I can without going over. Out on the edge you see all kinds of things you can't see from the center» Kurt Vonnegut.

newton.saber24-Oct-15 9:15

24-Oct-15 9:15

Good points. I agree.
The "how long it has been since author logged in" could just be a weight value or something.
Thanks.

Chris Maunder24-Oct-15 10:00

24-Oct-15 10:00

I read every suggestion, and they all get sorted and either fixed or put on the TODO.

The reason for the featured articles is to provide a place to pop up Really Good Stuff that you not otherwise catch. Changing it to restrict it based on author activity changes it to a system that would Show Active Authors Their Articles. Which is not a bad thing, but just not the original intent.

cheers
Chris Maunder

newton.saber24-Oct-15 15:07