|
may be he was talking to Chris and he did not wanted to post the problem in the forum for everyone to read,
Probably Chris will be interested no matter how it comes.
-Prakash
|
|
|
|
|
Can you do anything about accounts used only for advertising of commercial products, thus spamming a forum?
Like this one:
http://www.codeproject.com/script/comments/all_comments.asp?mode=all&userid=2404033
I didn't bother the first time but the company seems determined to repost their ads every couple of months. I understand that companies need to advertise their products. But as long as I don't post in their forums saying "hey guys, you can have the same for free at codeproject" I expect them to stay out of article forums unless they have something to say related to the article.
I guess the next time I will download their evaluation product benchmark it against mine and publish the results with full details and on the article. This could indeed be fun.
Regards
Ingo
|
|
|
|
|
Any spam is deleted as soon as it's reported.
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
What's up with searching message boards? It's painfully slow leading to non-use, leading to re-posts etc.
I just searched for the phrase "tape backup" in the lounge within the first four months of this year and it took as long as it took to type this message out to find it.
I can't imagine why it's taking more than 1 second for that type of search unless the search is implemented as some sort of brute force searching through the actual messages?
If that's the case, why not break and index every message as it's posted or modified into a dictionary and index table in the database and search that table instead?
This would reduce the search time by many orders of magnitude.
|
|
|
|
|
John Cardinal wrote: What's up with searching message boards? It's painfully slow
Yeah. It has never returned me any results. Just strugling and hang at certain percentage of the progress bar. Some one told me that it works --- just a little slow. But I waited for 30 minutes and got nothing. 
Maxwell Chen
|
|
|
|
|
I think they are using the Microsoft full text indexing built into SQL server. It's never been that fast, so on a site like this with competing queries it isn't going to be that zippy. Also if you hit it while it is rebuilding indexes it is going to take even longer.
|
|
|
|
|
Hmmm...I don't know about full text indexing, we considered it for a project and decided it wasn't good enough. Still I can't imagine it's that slow. I could do a brute force search in the time it takes to search even a few months in the lounge now so if it's not a brute force method it's badly broken.
I've built indexing and searching functionality in our business apps for years, it's pretty simple once you get the word breaking and excerpt retrieving down pat (even easier if you use my articles I published here on those exact topics  ).
To search a dictionary of unique words and cross reference them to an index of posts should be a split second operation.
I think it's a very important area to address because it's leading to all sorts of redundant posts and seriously limiting the basic functionality of this site.
|
|
|
|
|
John Cardinal wrote: To search a dictionary of unique words and cross reference them to an index of posts should be a split second operation.
It probably is, you just have to wait for all the other thousands of split second operations ahead of you, the database is handling thousands of queries per second. Plus there is probably a bit of memory pressure so unless something is used constantly the caches are going to be pushed out of memory and paged out to disk. Plus if there is one other common slow operation then under load everything will be pushed toward the performance of that query.
I'm not a fan of the microsoft search tools. Both from performance and user friendlyness. They should have a query mode like google that lets you use any type of query instead of telling it that you want to use boolean for example. I've had to use the integrated SQL engine in the past and ended up exporting documents out of the database and using index server to index HTML pages because we could get more control over how documents were searched and get better performance searching what would be multiple tables. Plus at the time the indexes had to be manually rebuilt. With index server the indexes were build incrementally as soon as a document was added. Plus getting the full text queries off SQL server helped out with performance.
John Cardinal wrote: I think it's a very important area to address because it's leading to all sorts of redundant posts and seriously limiting the basic functionality of this site.
I don't think there is going to be a good search until CP moves to a new technology for search or Microsoft productizes the stuff they are doing for MSN.
I can imagine the sinking feeling one would have after ordering my book,
only to find a laughably ridiculous theory with demented logic once the book arrives - Mark McCutcheon
|
|
|
|
|
I understand the load issues, but rolling your own word breaking, indexing and searching stuff isn't rocket science and it's quite easy to squeeze many orders of magnitude better performance out of that when it's custom built for a specific application than a generic multi-purpose app.
Searching is such an important feature that if it's truly a load issue that's hampering it, it really should be a dedicated server for indexing and searching.
I don't know, it's obviously more difficult than it seems from the outside and running a website for developers can't be easy with a million of us in the peanut gallery thinking how we could do it better. I guess I should just accept it and worry about my own stuff! 
|
|
|
|
|
John Cardinal wrote: I don't know, it's obviously more difficult than it seems from the outside and running a website for developers can't be easy with a million of us in the peanut gallery thinking how we could do it better. I guess I should just accept it and worry about my own stuff!
I'd imagine that is the hardest part, especially since we all have our own little pet peeves, search, performance, HTML layout, etc. Though, better forum search is up there with things I'd like to see. At least most of the 500s and timeouts have cleared up.
I can imagine the sinking feeling one would have after ordering my book,
only to find a laughably ridiculous theory with demented logic once the book arrives - Mark McCutcheon
|
|
|
|
|
This is a priority. Just give me a few days to recover from jetlag and put out any fires that started during my two weeks off 
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
Just increase the date gap. The first one. It should be above 3(which is the default) months. Then it searches like a tracer bullet.  Make it something like 5 months or more and see the results.
Nibu thomas
Software Developer
Faqs by Michael dunn
|
|
|
|
|
Nibu thomas wrote: Just increase the date gap. The first one. I should be above 3 months. Then it searches like a tracer bullet.
You meant we have to initialize the search function?!
Maxwell Chen
|
|
|
|
|
Maxwell Chen wrote: You meant we have to initialize the search function?!
Hehe. Helpless had to do it. As it was taking years to search. So I changed the parameters in the first one to check. And Lol Lol Lol Lol I couldn't believe it. ![Jig | [Dance]](https://codeproject.global.ssl.fastly.net/script/Forums/Images/jig.gif)
Nibu thomas
Software Developer
Faqs by Michael dunn
|
|
|
|
|
Cool !! 
Maxwell Chen
|
|
|
|
|
|
I knewly submitted an article, displayed on the main page in the Latest section...but not in its RSS feed 
what is the problem?
|
|
|
|
|
The RSS feeds are split into categories and only display the latest 10 articles over a certain rating threshold. It could be you were looking at a feed which didn't include your article, or your article had been pushed out by a newer article (and the homepage cache hadn't been updated yet) or your article fell below the rating threshold
cheers,
Chris Maunder
CodeProject.com : C++ MVP
|
|
|
|
|
Someone must have mentioned it before already, but it would be nice to receive notifications if an article of interest gets updated. It could be added as a checkbox for every bookmarked article in the "My Bookmarks" page or it could be an option on its own.
Thanks
Damir
|
|
|
|
|
|
 
And still bug with "onclick=" in posts ("onclick=")
Fo more info see this topic
Best regards, Alexey.
|
|
|
|
|
Mihai Moga wrote: Hi, some of the author's list articles are still buggy e.g. Articles by Jens Schacherl[^] Could you please fix that? 10x
What's the bug there? The page loaded fine for me just now.
Regards,
Nish
|
|
|
|
|
The problem is - all articles show up in a single category (".NET/Active Directory")...
Please fix.
-- modified at 5:23 Wednesday 3rd May, 2006
Fixed, thank you.
|
|
|
|
|
Hi ,
Why not have a separate core C++ forum ? instead of messing the things with VC++ UI and Win32 forums?
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
