|
Hi,
While answering this question:
Need help on printf function[^]
All hell broke loose with the javascript editor... it was actually modifying my c++ code... it changed the line:
vector<unsigned char> v;
into:
vector<unsigned char = ""> v;
Then it added some wierd tag... </unsigned>
Thanks,
-David Delaune
|
|
|
|
|
This is a known bug where the editor closes open tags on occassion.
You effectively opened a tag <unsigned>.
You need to ensure you use > and < to encode the <>.
|
|
|
|
|
Well,
As a user of the website... basically I just want it to work. If it does not work then it is obviously broken and I will report it. I am certainly not an expert with HTML/Javascript but the site should probably avoid using Tags or META Tags that are the same as C#,C++ and VB language keywords[^].
Best Wishes,
-David Delaune
modified 31-Dec-11 8:12am.
|
|
|
|
|
Randor wrote: As a user of the website... basically I just want it to work.
Me too, I wasn't saying it won't be addressed, all I was doing was confirming that your bug is known about and on the hamsters list of things to fix 
|
|
|
|
|
DaveAuld wrote: your bug is known about and on the hamsters list of things to fix
I have a feeling... that it is a very very long list. Perhaps even be longer than Santa's naughty and nice list.
-David Delaune
|
|
|
|
|
It's not buggy - it just doesn't work as you expect.
By default the editor accepts HTML, so if you type in "vector<unsigned char> v;" then it thinks the < is the start of a HTML tag and processes it as if it were HTML: namely, it tries to ensure it's well formed HTML (hence the weird closing tag), then colourises anything it needs to colourise, then hands it back to the browser for rendering. The browser has no idea what a "unsigned char" tag is so does the correct thing (according to it) and ignores it.
If you had placed the code in a PRE block then our pre-processor / HTML Tidy-up code will auto-encode the < entity and so you would get
vector<unsigned char> v;
When pasting text we sniff what you've pasted and try to guess what's in the text. If we think we see code we'll wrap it in a PRE block (so you get the auto-encode) otherwise we leave it. In either case the Paste Preview window pops up showing you exactly what your pasted text will look like when rendered based on how you choose to paste the text.
You can disable the HTML processing by checking the "Treat my content as plain text, not as HTML" checkbox just below the message preview.
Our options in all of this are:
1. Don't accept HTML at all.
Not a great option because we like being able to have formatted text, and we like being able to wrap code in PRE blocks and have it colourised.
2. Make the editor even smarter.
I'm open to suggestions. The most obvious improvement is that we specifically check all <tag> blocks to see if it is a true HTML tag, and if not, encode the entities. However, this will totally fail for any HTML code posted.
3. Switch to MarkDown.
While this would help with some things, but not others. You still have to indent your code, for instance, 4 spaces to have it formatted correctly, as opposed to wrapping in a PRE or CODE block.
4. Switch to a pure WYSIWYG editor
We tried this. Everyone hated it. WYSIWYG editors depend on the underlying browser editor, and they are all different, and they are all painful to use and to bend into shape to make them usable.
I'm open for further suggestions on how to make this more intuitive.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
My suggestion: leave it as is, it's a very small learning curve and really quite simple to use.
Unrequited desire is character building. OriginalGriff
I'm sitting here giving you a standing ovation - Len Goodman
|
|
|
|
|
Richard MacCutchan wrote: it's a very small learning curve and really quite simple to use.
Richard, that is easy to say for someone spending 10% of his day and 10% of every week on the website. I can also say that every software product of which I was involved in the design process... was completely intuitive to me. What was really important was how intuitive the interface was for new or infrequent users.
I personally believe there is *always* room for improvement in every piece of software. Some aspects of the codeproject website that I think could be improved:
1.) The javascript editor should probably be smarter and make an attempt to check for language keywords masquerading as html/meta tags. All of the keywords are documented and standard... this seems like an obvious improvement considering the codeproject forums will contain a large amount of code.
2.) The 'Quick Answers' comment button labeled 'Have a Question or Comment?' seems to be missed by new users very frequently for some reason. They often answer their own question as a comment. It is easy to call the new users 'morons' but I think some of this can be blamed on the UI.
That's all I can really think of for now. Overall I really like the new features with the exception of the voting system.
Best Wishes,
-David Delaune
|
|
|
|
|
Randor wrote: that is easy to say for someone spending 10% of his day
A valid point, it's now a few years since I first started using it.
Randor wrote: there is *always* room for improvement
Again very true, although I do take the view that the cost of CodeProject is such that it is difficult (for me) to justify too much criticism.
Unrequited desire is character building. OriginalGriff
I'm sitting here giving you a standing ovation - Len Goodman
|
|
|
|
|
Chris Maunder wrote: If you had placed the code in a PRE block then our pre-processor / HTML Tidy-up code will auto-encode the < entity and so you would get
Chris,
I did have the code wrapped within a <pre lang="c++"> tag. In fact... I am testing it right now and cannot get the behavior you are describing when pasting as Text. In other words... even if I have the PRE tag its modifying the c++ code.
Should I be pasting as 'Text + encoded HTML'? The preview looks correct if I paste with this format.
Best Wishes,
-David Delaune
|
|
|
|
|
Since I've written 3 articles in the past couple of weeks, I've found the easiest route to authoring articles is the following:
* Choose a markdown editor of your choice (I use MarkdownPad).
* Write the article in markdown on your own PC.
* Export as HTML, then paste everything inside the body tags into the CodeProject editor
* Add the finishing touches, and adjust image/zip links as necessary.
* Have a cup of tea.
I can imagine that the challenge of using Markdown as the method of submission is a huge challenge on the back-end of CodeProject, especially when regenerating markdown for article updates. Markdown would be nice, but as you say, it's not as rich as HTML.
You could always add some checks for the usual suspect keywords such as map, vector, list, shared_ptr and so on followed with a < , to at least give the article write some cursory warning of a potential problem. Or do the reverse: have a list of all HTML elements: p, table, thead, img, etc. and perhaps pattern match against them, and anything that doesn't match, flag up and warn.
Regards,
Barry
|
|
|
|
|
Barry Lapthorn wrote: Or do the reverse: have a list of all HTML elements: p, table, thead, img, etc. and perhaps pattern match against them, and anything that doesn't match, flag up and warn.
That's essentially what I've done - at least within PRE blocks. Anything that smell like a HTML tag that isn't among the anointed gets encoded.
Converting from HTML to Markdown wouldn't necessarily be an issue because we'd simply keep the back-end text as the original markdown. It would mean no WYSIWYG editor if you chose Markdown, but I don't think those who used Markdown would care (as long as the editor provided shortcuts and helper menus)
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Chris Maunder wrote: Converting from HTML to Markdown wouldn't necessarily be an issue because we'd simply keep the back-end text as the original markdown.
@chris: that's good to know - I've written some CP articles recently using markdown and the steps are: markdown, export to HTML, post to CP, then *add language type* to syntax highlight code snippets: meaning that you can't author a CP article purely in markdown if you want to have code snippets syntactically highlighted (I think!!). I know markdown supports embedded html, but I think what you've already done works sufficiently 
Regards,
Barry
|
|
|
|
|
Basically this is my laziness to move my hands from keyboard to mouse (or use two hands to press Alt+F4), but it would be nice if the copy code window which opens from a pre-tagged code block could be closed by pressing Esc.
|
|
|
|
|
See if we can provide list of members who viewed the question.
|
|
|
|
|
Now what would be the benefit of that?
- I don't see any sense in a feature such as this.
- Why would anybody want to be this nosy?
- Could be easily defeated by browsing CP anonymously.
Manfred
"With sufficient thrust, pigs fly just fine."
Ross Callon, The Twelve Networking Truths, RFC1925
|
|
|
|
|
Rahul Dhoble wrote: See if we can provide list of members who viewed the question.
You want to see the list with or without paging?
If you want to see the questions by type, here you go Popular[^], Top Rated[^]
|
|
|
|
|
I know it may a very big list and just checking if we can add this cool option in our portal
|
|
|
|
|
You still haven't mentioned why anybody would want to know this.
What is to be gained by a feature such as this and why would I want you to know if I read a question.
All I can see is this unfortunate scenario:
- OP posts question
- OP returns later to see who viewed his question
- OP pesters everybody on the that list in order to garner some support.
- Some tortured souls who had answered to OP's question go through the list and pester everybody who only viewed that question with even more obnoxios requests to have a look at their solutions. (Still the SAme I shall drop no name!)
Cheers!
Manfred
"With sufficient thrust, pigs fly just fine."
Ross Callon, The Twelve Networking Truths, RFC1925
modified 30-Dec-11 10:35am.
|
|
|
|
|
I'm Sure Anybody Kould read your unnamed kall out very easily... 
Ideological Purity is no substitute for being able to stick your thumb down a pipe to stop the water
|
|
|
|
|
|
No, Actually the 1st line in my message is sarcastic one, Yes I agree with Manfred.
|
|
|
|
|
I can't see how anyone would benefit from this: for example, I view many questions I do not answer because the tags are missing, or I didn't read them correctly, and the question is on a subject I am not familiar with. How would it help anyone to know that I viewed your question?
Ideological Purity is no substitute for being able to stick your thumb down a pipe to stop the water
|
|
|
|
|
Are you and dr.kitanjal[^] the same person?
The posting behaviour (of solutions) to this question[^] looked odd, one answer deleted under one user account and resubmitted under another user account.
Then looking back at some messages from a couple of weeks back there was activity from around the same time and same sort of daft questions being asked.
I have my suspicions....
|
|
|
|
|
| No
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
(that "kall out" part had me spurting my 