|
Can you please send me a screen shot?
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
As you wish so it shall be.
Jon
'When once a republic is corrupted, there is no possibility of remedying any of the growing evils but by removing the corruption and restoring its lost principles; every other correction is either useless or a new evil.' ~ Montesquieu
Soap Box 1.0: the first, the original, reborn troll-less
|
|
|
|
|
It should be better now. working on a more robust solution.
Robust == nasty workaround for IE7 and below's crap rendering.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Chris Maunder wrote: It should be better now.
Nope. Same effect. May I suggest that rather than putting the break after the "small text" class you put it before? by the way, I am using I.E. 8.0.6
Jon
'When once a republic is corrupted, there is no possibility of remedying any of the growing evils but by removing the corruption and restoring its lost principles; every other correction is either useless or a new evil.' ~ Montesquieu
Soap Box 1.0: the first, the original, reborn troll-less
|
|
|
|
|
My diff algorithm is returning "The lines are identical"
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
yep I couldn't get it formatted properly, so I decided to use English on the edit.
Jon
'When once a republic is corrupted, there is no possibility of remedying any of the growing evils but by removing the corruption and restoring its lost principles; every other correction is either useless or a new evil.' ~ Montesquieu
Soap Box 1.0: the first, the original, reborn troll-less
|
|
|
|
|
Actually I uploaded the wrong CSS file. Can you try now?
The correct fix (for that situation) is that the tabs should be inline-block, not inline. However, that is causing the selected tab to disappear on another page in an odd way.
IE8 standards + everything else works fine
IE7 and below: broken.
As a workaround I will insert a div with clear:both for IE7--
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Works fine for me.
Chris Maunder wrote: As a workaround I will insert a div with clear:both for IE7--
Will that fix it for FF2 as well? (IIRC, it doesn't support inline-block either)
Maybe you should just add a popup saying that if you're still using an old browser, you don't deserve to have a private group.
Jon
'When once a republic is corrupted, there is no possibility of remedying any of the growing evils but by removing the corruption and restoring its lost principles; every other correction is either useless or a new evil.' ~ Montesquieu
Soap Box 1.0: the first, the original, reborn troll-less
|
|
|
|
|
As it is monday, lets keep it a simple one (actually 3 bugs):
By default the syntax colorizer recognizes strings literals by looking for double quotes, and it knows \" isn't a delimiter, however it goes astray on a simple line such as string s="This is \\"+"bad";
which yields
[1] s="This is "+"good";
[2] s="This is \\"+"bad";
[3] s="This is bad too";
bug 1:
a double backslash is not interpreted correctly: An even number of escapes doesn't affect the following character. The correct rule would be:
Double quotes are delimiters unless preceded by an odd number of consecutive backslashes.
Similar things may apply to other languages and other escape characters (just in case you decide to add a Regex forum, and LANG=REGEX some day).
bug 2:
the "inside string" state is carried from line 2 to line 3; C# does not do that (C and C++ accept this IIRC), each line needs its string literal closed correctly.
bug 3:
the real example was in one of the programming forums today (C#? NET Framework?) however I couldn't find it again, as search for \\" simply doesn't work at all. There are times full text search based on word indexing isn't good enough.
Luc Pattyn [Forum Guidelines] [My Articles]
The quality and detail of your question reflects on the effectiveness of the help you are likely to get.
Show formatted code inside PRE tags, and give clear symptoms when describing a problem.
|
|
|
|
|
Yes, unfortunately this is an old bug that has driven me nuts but I'm going to have one of the other guys with more patience than me dig in. It's a RegEx issue.
BTW: was it you who wanted the <pre linecount="on"> option? How's this look?
1 Line
2 Another line
3 A further line
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Chris Maunder wrote: was it you who...
yep. I'm guilty of most readability suggestions, coloring remarks, etc.
Chris Maunder wrote: How's this look?
1 <<<< I'd rate it here, lacking anything higher
excellent.
Can we get a user setting to have it ON all the time (say a fifth checkbox, with persistence)?
And you might turn it ON by default, otherwise it will take ages before the majority starts using it.
While you're at it, you might consider a selectable default language, mine would be 'cs' for PRE
and 'none' or 'text' for CODE (that's if you didn't throw coloring out all together for CODE blocks).
|
|
|
|
|
Three more ideas on PRE blocks, actually on Collapse code snippet PRE blocks in articles (and probably also applicable in forums):
1. It isn't always code that goes into a PRE block, it could be an input file, an output file, in fact anything that can take advantage of a non-prop font and a colored background. So the text isn't 100% OK.
2. It would be nice to get the PRE blocks themselves numbered automatically. So 1+2 could result in:
Collapse block #1
With numbered blocks you may avoid the request for an initial line number to differentiate line 1 in one PRE block from line 1 in another PRE block.
A problem now is text will start to refer to such blocks, then upon editing and inserting/removing a PRE block, the references may become wrong. Not sure how to solve this.
[EDIT]Maybe accept a name='text' attribute in a PRE tag, and show the name, when present, instead of the sequential number.[/EDIT]
3. It doesn't make much sense to collapse a one-line block, so for those I would reduce it to:
block #1 .
I remember having seen an article-needing-approval that consisted entirely of code, each line in a separate PRE block, all collapsible and the whole thing looking horrible.
|
|
|
|
|
Definitely #3, but I'm not sure how much #2 would be used. It's a nice features, definitely, but I'll leave this on the back-burner for now.
"name=text" is also a nice idea but only you and Hans would use it. I fear this is a bigger issue and one that needs a simpler, automatic solution.
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
The 20 articles I have written are shown on my MemberArticles[^] page, but the count at the top says 21.
Also, the Who's Who list, sorted by Article Count, seems to be messed up. For example on the first page, number 23 is 'Jijo raj' with 1 article! And if you go to his MemberArticles[^] page, it says he has 1 article, but no details are listed.
Thanks,
Nick
----------------------------------
Be excellent to each other
|
|
|
|
|
Seems this bug is persistent
I'v reported it a couple months ago as well, and someone else did a couple month's / weeks before that
|
|
|
|
|
A bit of debris from this weekend's construction.
Just let me find a broom...
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Pick any one
Luc Pattyn [Forum Guidelines] [My Articles]
The quality and detail of your question reflects on the effectiveness of the help you are likely to get.
Show formatted code inside PRE tags, and give clear symptoms when describing a problem.
|
|
|
|
|
As of this writing, I haven't published any articles at CP, but my profile shows that I've submitted 7 articles.
Thanks Chris. Like I promised, I'll buy you a few beers.
It is a crappy thing, but it's life -^ Carlo Pallini
|
|
|
|
|
This is a bug. Just not a bug in the way you think it is. You do actually have 7 articles.
Remember: there is no spoon...
cheers,
Chris Maunder
The Code Project Co-founder
Microsoft C++ MVP
|
|
|
|
|
Stop... You make me think if I should have taken the blue pill.
It is a crappy thing, but it's life -^ Carlo Pallini
|
|
|
|
|
Chris Maunder wrote: there is no spoon...
Maybe you could for once drink the beer without using a spoon...
Luc Pattyn [Forum Guidelines] [My Articles]
The quality and detail of your question reflects on the effectiveness of the help you are likely to get.
Show formatted code inside PRE tags, and give clear symptoms when describing a problem.
|
|
|
|
|
|
I noticed that on my profile page, I am credited with submitting 4 articles - odd, since I haven't submitted any! But that's not too bad, my average rating is 3.92 with an empty list...
No trees were harmed in the sending of this message; however, a significant number of electrons were slightly inconvenienced.
This message is made of fully recyclable Zeros and Ones
|
|
|
|
|
I see that my average article rating is better than yours.
BTW, I found the bug myself (before reading your post).
It is a crappy thing, but it's life -^ Carlo Pallini
|
|
|
|
|
For some 10 minutes a message in C# forum is consistenly present when I happen to be served by web13, and absent otherwise (web09, web20). Trying to add a reply fails with a "Cannot find message" which gets shown also by web13.
The message by new member Allan R Landall evolved from "just posted" to "10 mins ago". I don't have its subject line or content, it was about paint events, and fitted above this one: http://www.codeproject.com/Messages/3170195/Csharp-Class-Library-How-do-I-make-my-class-interop-and-registerable.aspx)
for the last two months there weren't hardly any Sundays the site worked properly; the symptoms aren't always identical, the net effect is posting replies does not work, and signaling such on sugs/bugs forum doesn't work either. I would like to see this problem fixed once and for all, it is very annoying and does not reflect well on our community.
BTW1: the "Cannot find message" error page doesn't allow simple page changes, it is protected as if it holds dirty data, which it doesn't.
BTW2: Here is the reply I wanted to post (I had in on clipboard as always):
Hi,
welcome to CodeProject. Hope you'll enjoy it.
you can signal a Control needs a repaint by calling its Invalidate() method.
Allan R Landall wrote: ... is in the Forms Paint Event, because that methods PaintEventArgs makes this so much easier
No, the primary reason for drawing it all in the paint handler is that it will be invoked automatically when other system activity makes that necessary, as in form minimize-restore, form maximize-restore, form getting uncovered by some other window, etc. And yes PaintEventArgs offering a Graphics object is a big plus.
Luc Pattyn [Forum Guidelines] [My Articles]
The quality and detail of your question reflects on the effectiveness of the help you are likely to get.
Show formatted code inside PRE tags, and give clear symptoms when describing a problem.
|
|
|
|
|