|
Browsing the CP Who's Who, ordered by Article count, the first page took a long time (ca 15 seconds) to show up, then yielded 25 people, starting with Marc Clifton (135 articles), and ending on razesh.kolla@gmail.com (0 articles). In fact the last three showed 0 articles, so that would imply a total of less than 2000 articles.
The second page starts with Uwe Keim (38 articles) and ends with Red Gate Software (1 article).
Page 3 starts with Roger Allen (26 articles), etc.
Something definitely is wrong, nothing serious, but wrong it is.
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
Local announcement (Antwerp region): Lange Wapper? Neen!
|
|
|
|
|
This is partly a load issue (the time) and partly a bug (the article counts).
It's actually good timing because we've been discussing the specific manner in which we count articles for members and have been looking to denormalise our tables a little in order to help with performance in the case of sorting by article count. The debate was whether or not there is currently an issue with sorting by article count. We never saw it ourselves, but clearly it's an issue. Discussion ended.
As to the counts themselves, this is easily fixed.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Thanks.
Not sure I understand all this, according to the home page there's some 26K articles.
If they were all in one table, I would expect grouping and sorting that could be almost immediate. I guess having multiple authors (and maybe groups) complicates things a bit.
Anyway, I don't mind waiting a few seconds for this info, as I don't need it frequently; I do hope we always get correct data however.
FWIW: I would be very interested in reading a couple of articles on the CP database stuff itself, how things are organized: the infrastructure, maintainance, DB design, code structure, ... It is after all a big application we are familiar with as a user, and could learn from a lot.
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
Local announcement (Antwerp region): Lange Wapper? Neen!
|
|
|
|
|
Counting articles is easy. Sorting 6.5 million members joined to a table 26,000 rows long takes time. You can always pre-aggregate the article table and then join these results with the member table and then sort on that, but that reduces the join from 6.5M on 26K to 6.5M on around 9K and so it doesn't save you much, and in fact means you have a massive temp table. So we'll add a new "ArticleCount' column to the user's table and ensure it stays in sync with reality. Kinda yuk, but the speed improvement will be massive because we can then put an index on that column.
As to the articles we've been tossing around a few ideas on exactly this - though possibly in podcast form where it's more of a thing where we sit down and you get to hear how we work through issues (and also get to hear about the bizarre and odd issues that you guys never see).
I like articles, but a podcast is something that can be created while we're actually working.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Chris Maunder wrote: "ArticleCount' column
I understand now, thanks. I hope you will use the new ArticleCount column only in situations where it doesn't really have to be real-time exact (as in listing users by article count); I would hope to see the real value when looking at my own page, say while I am in the process of adding/editing/removing articles and tips&tricks and what-have-you.
Chris Maunder wrote: podcast
looking forward to that.
Chris Maunder wrote: bizarre and odd issues that you guys never see
we sometimes do get our share of bizarre and odd results though. 
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
Local announcement (Antwerp region): Lange Wapper? Neen!
|
|
|
|
|
It will be an up-to-date value, always.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Thanks Chris, and Elina.
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
|
|
|
|
|
When you post a message, the 'new' image doesn't appear (which is  ), but the menu link for the forum which you just posted in is still bold, even though no-one elses posts are new.
If at first you don't succeed, you're not Chuck Norris.
|
|
|
|
|
Checking each forum for whether or not *you* have a new message in there is too much load. Sorry 
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
In this[^] post, the color formatting has a couple of issues, such as strings not closing properly and comments not in a different color.
Also, I had to change the < to < (and > ) manually for the links and the divs, even though they were within pre tags.
If at first you don't succeed, you're not Chuck Norris.
|
|
|
|
|
I agree that the syntax colorizer has problems (I've noted that problem in the past). However, the pre tag is a different story...
The pre tag is not supposed to escape inequality symbols. It preserves spaces and line breaks. To encode inequality symbols automatically, click the "Encode HTML tags when pasting" option below new messages.
There will be a line break after this sentence because pre does not escape inequality symbols.<br />See?
If I select the aformentioned option to encode HTML tags, copy the above pre tag in its entirety (i.e., including its content), then paste it and surround it with pre tags again, I get the following:
<pre lang="text">There will be a line break after this sentence because pre does not escape inequality symbols.<br />See?</pre>
Visual Studio is an excellent GUIIDE.
|
|
|
|
|
I believe you missed my point.
pre tags are used on this site to show code in messages. My point is that if you paste code directly into the pre tags (such as html) then the editor should automatically escape it, without the need to select "Do not interpret HTML tags (good for code snippets)". Like when a url is pasted into the editor it automatically adds the link etc.
If at first you don't succeed, you're not Chuck Norris.
|
|
|
|
|
I tend to agree with you, for several reasons:
1. it would be more comfortable when code can simply be copied in VS and pasted in CP without having to worry about some checkboxes hidden underneath the edit window;
2. it would reduce the number of unreadable messages, where people did it wrong and never notice; a lot of questions about generics are completely lost right now;
3. it would be more consistent with other editors and HTML viewers.
The official reply seems to be people want the ability to have some HTML formatting going on inside PRE blocks; as an example, <b> is often considered useful to highlight the line where something goes wrong.
Therefore my suggestion has been to get automatic escaping of whatever gets pasted inside PRE tags, leaving the possibility to apply internal HTML tags once everything got pasted in. As a refinement, the automatic escape logic could (optionally) skip a limited set of tags, such as <b>
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
Local announcement (Antwerp region): Lange Wapper? Neen!
|
|
|
|
|
What if we simply used a WYSIWYG editor?
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Can you expand on that?
If you are referring to the article submission wizard, I had a hell of a time getting my TextFileDiff article look the way I wanted; it has some PRE blocks showing code and others showing app output (which includes HTML stuff from your test files). It took me 4 iterations to get all the < signs right in both code and HTML output; times two, as I did one revision which started with removing everything and pasting my source again. (And I should do another revision, but am not looking forward to that right now).
In my experience WYSIWYG is never the full answer.
Just like a GUI-only app isn't complete (I tend to need command line "extensions" to WinZip, which are available as a separate install!)
We always need some copy-and-paste facilities, making copying code from IDE to publication (and vice versa) easy.
And when formatting capabilities are present, they should be encoded somehow, so we can copy the full source (text+formatting) in and out too (I would accept explicit export and import functions for that).
So what is it you have in mind?
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
Local announcement (Antwerp region): Lange Wapper? Neen!
|
|
|
|
|
The article submission wizard and then the new Quick Answers system.
Luc Pattyn wrote: It took me 4 iterations to get all the < signs right in both code and HTML output
Were you in Design mode or raw HTML mode? Can you give me an example of the type of text/code you were pasting and what you were trying to achieve?
Everyone does everything slightly differently and the WYSIWYG editor works really, really well for me, but obviously because I wrestled it into submission for myself. Then again, I never use Winzip from a command line (ah, those were the days...)
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
The examples are in here[^].
I copied the lot from my HTML file (which looked exactly the way I wanted it in FF3, given a preamble with some elementary styles) and pasted it into the wizards editor. Don't recall what mode it started of in; it took me a while to discover it has two modes, it does not tell, nor does it tell which mode it is in. I do recall it behaved differently for code (with an occasional < in a for loop) and HTML-tagged text data; it felt like a bug but I never was able to put my finger on it, otherwise I would have reported it.
BTW1: why is it the WYSIWYG editor has a lot of functionality that becomes unavailable in source mode? e.g. I would like to be able to search for something in the source. And a find-and-replace would have eased things a bit (so I could turn all & into & or the other way around).
BTW2: the current Q&A seems to have three modes: two in Edit, one in Preview. Not sure I understand that either. Is preview identical to WYSIWYG, except it also is read-only? If it were really WYSIWYG why would I still need a preview?
PS: I need command line interfaces for automation, e.g. batch files creating ZIP files. That is how I create the download packages for my articles, and much more.
Luc Pattyn
I only read code that is properly indented, and rendered in a non-proportional font; hint: use PRE tags in forum messages
Local announcement (Antwerp region): Lange Wapper? Neen!
|
|
|
|
|
Luc Pattyn wrote: it took me a while to discover it has two modes, it does not tell, nor does it tell which mode it is in
In design mode you are editing WYSIWYG. It should be very, very obvious that you have formatted text, images, layout etc. In source mode you see plain text in Courier New and lots of HTML tags. It's the usual HTML mess.
BTW1: Some things we get for free in WYSIWYG mode because browsers give it to us. Some of these things are hard to implement within HTML mode. Some things we simply haven't implemented.
BTW2: In Q&A yes, there is WYSIWYG, Edit and preview. Preview gives you a final sanity check and will colourise code snippets.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
In forums that are being kept private, you can see the number of pages (and you can go to those pages, but of course they show as empty since they are private)
I'm not sure whether that's really a bug, but it seems odd to me..
|
|
|
|
|
Should be fixed.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
I attempted to paste some text (copied from notepad) into a new message on this site bugs forum and the text was placed inside of an HTML anchor tag as the HREF. I think the problem was that the start of my text was a URL and the algorithm that detects URL's thought the entire block of text was a URL (newlines, spaces, punctuation, and all).
Visual Studio is an excellent GUIIDE.
|
|
|
|
|
Can you re-paste the message so I can have some testing data?
Thanks!
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
|
Also, after I paste a URL, the text inside the anchor tag typically gets highlighted. While that happened with the above test message, the highlight was off by a few characters, likely due to the newline characters in the message I pasted. FYI, I am on IE8 and compatibility mode is NOT turned on. Windows XP Pro SP3.
Visual Studio is an excellent GUIIDE.
|
|
|
|
|
Fixed. Will uplload soon
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
