User talk:Remember the dot/Archive/2012
This is an archive of past discussions with User:Remember the dot. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Documentation for gadget authors
We're trying to start a library for gadget authors to use. Please check it out and post any questions or comments there. -- ☠MarkAHershberger☢(talk)☣ —Preceding undated comment added 01:36, 9 March 2012
Your syntax highlighter
I just popped your syntax highlighter into my common.js after seeing you mention it on the Signpost discussion. I haven't done any "real editing" with it yet (that will be coming...), but for now I just want to say thank you thank you thank you! I've been looking for something like this recently, and your tool seems to do exactly what I need it to do, no more and no less. A fluffernutter is a sandwich! (talk) 00:59, 26 August 2012 (UTC)
- Haha I am very happy you like it! Tell your friends! I really like it too and I was disappointed that no one else seemed to care. Let me know your thoughts after you've been using it for a while. —Remember the dot (talk) 05:11, 26 August 2012 (UTC)
Indeed, this is fantastic! I've left some detailed notes and questions at WP:VPT#Syntax highlighter reboot (where I saw it), and linked to that thread from WP:VPR#Differentiating reference syntax in the editing window. Best wishes, and a very hearty "welcome back! ye were missed." :) -- Quiddity (talk) 22:23, 28 August 2012 (UTC)
People care, but Wikipedia:Village pump (technical) is not the best place anymore to get interest. Also, people do not have the ability to watchlist individual discussions. See:
- Wikipedia:Village pump (proposals)#Proposal for one Village Pump using DPLforum or other subpage discussion software.
- Wikipedia:Village pump (proposals)#Support or oppose a test Village Pump based on Portuguese VP.
I suggest you formally propose this as a gadget at Wikipedia:Village pump (proposals). Then you will definitely get interest. --Timeshifter (talk) 03:15, 29 August 2012 (UTC)
I think the script is a great idea, but it doesn't seem to work on Google Chrome. A boat that can float! (watch me float!) 11:31, 30 August 2012 (UTC)
- Thanks for trying it. It's problems like these that need to get worked out. When you say "it doesn't seem to work", what do you mean? What happens when you try to use the script? What is your User-Agent? Does it work for you in Firefox? —Remember the dot (talk) 22:38, 30 August 2012 (UTC)
- Whoops, sorry it took so long to reply. The truth is that, well, nothing happens when I use it. All the text stays black. A boat that can float! (watch me float!) 06:27, 1 September 2012 (UTC)
- It works fine here...I need your User-Agent information to be able to figure out what's going on on your end. —Remember the dot (talk) 13:24, 1 September 2012 (UTC)
- Finally got back to the computer I was using when I installed it. My user agent is Mozilla/5.0 (Windows NT 6.0) AppleWebKit/537.1 (KHTML, like Gecko) Chrome/21.0.1180.83 Safari/537.1. A boat that can float! (watch me float!) 08:43, 3 September 2012 (UTC)
- I don't see anything out of the ordinary in your User-Agent, though I could be wrong. You said you've been using more than one computer? Have you tried Firefox instead of Chrome? If nothing you do seems to help then maybe you have a gadget installed that somehow conflicts with the syntax highlighter. —Remember the dot (talk) 23:12, 4 September 2012 (UTC)
- I tend to use about 3-4 computers regularly, including an iPad. I'll try Firefox as soon as I can. A boat that can float! (watch me float!) 12:48, 14 September 2012 (UTC)
- Just tried Firefox; still doesn't work. My user agent this time is Mozilla/5.0 (Windows NT 6.0; rv:15.0) Gecko/20100101 Firefox/15.0.1. A boat that can float! (watch me float!) 07:46, 15 September 2012 (UTC)
I have it installed for a few days now. I am going to uninstall it though. It is kind of distracting because it highlights so many things. If it highlighted only references, then I would keep it installed. That would be so useful. It wastes so much time to have to hunt through a lot of text to find one reference out of dozens. Especially when one reference is used multiple times, and only one of the locations for it in the text has the full reference details. Using find in the browser is helpful, but I like being able to just scan the page for references, too. That is where your gadget is so helpful. --Timeshifter (talk) 02:36, 2 September 2012 (UTC)
I hope you don't mind me posting here instead of the meta talk page. I rarely check my meta watchlist. I kept getting this message while adding a banner a little while ago to the top of many pages:
A script on this page may be busy, or it may have stopped responding. You can stop the script now, or you can continue to see if the script will complete. |
I removed the JS from User:Timeshifter/common.js. I was going to remove it anyway since I don't like so much being highlighted. I am looking forward to the custom version that only highlights references. --Timeshifter (talk) 07:15, 2 September 2012 (UTC)
- Can you explain in as much detail as possible what you were doing that caused the script to freeze? You seem to have been adding {{rank order}} to several pages but when I look at those pages nothing freezes. —Remember the dot (talk) 19:05, 2 September 2012 (UTC)
- I have been adding that banner to the top of many pages. I usually do 10 or 20 pages at a time. I go to a category I am working on, and I go in alphabetical order. I open up around 10 pages in 10 tabs. Then I quickly click the edit link for the lead section for each page. So the edit windows for the lead sections for 10 pages are all opening up around the same time. This must overwhelm the script and/or the PC. This type of rapid editing happens often on the Commons when adding or changing categories for multiple images. Especially for people who do not know of all the Commons gadgets. I do not know how often this happens on Wikipedia.
- Maybe you can add an an/off toggle button above the edit window. So the script does not operate again until the button is clicked again above an edit window. In other words when I know I will not be needing the highlighting I can turn it off for that day. I can turn it back on again another day or later in the day. --Timeshifter (talk) 04:20, 3 September 2012 (UTC)
- In the future the highlighter will be a gadget and you'll be able to turn it off through the Preferences menu if it slows your computer down too badly. —Remember the dot (talk) 23:12, 4 September 2012 (UTC)
Testing in Internet Exploder on Linux
If you have enough free disk space to store them and a computer capable of running QEMU/KVM, Microsoft provides freely downloadable VHD images that can be converted to work. Anomie⚔ 20:43, 2 September 2012 (UTC)
- Interesting, thanks for the link. Is there an IE9 or IE10 image available? (I suppose I could just take one of the images and update it.) —Remember the dot (talk) 22:36, 2 September 2012 (UTC)
- They've got a bloated IE9 image listed there, no IE10 yet. Anomie⚔ 01:53, 3 September 2012 (UTC)
- I tried the IE8 and IE9 images with VirtualBox but both gave me a "corrupted ACL" error and wouldn't boot. Have you had better luck? —Remember the dot (talk) 23:12, 4 September 2012 (UTC)
- I've been using the IE8 image for occasional testing for some time now with qemu (specifically,
qemu-system-x86_64 -enable-kvm -name 'Maleficium',process=maleficium -cpu host -m 1024 -usb -vga std -net nic,model=e1000,vlan=1,macaddr=52:54:00:12:01:03 -net vde,vlan=1,sock=/var/run/vde2/vde0.ctl -soundhw hda -hda Win7_IE8.vhd -snapshot
). Never tried VirtualBox. The only real annoyance is that it can't seem to deal with qemu's emulated touchpad, so it has to grab the mouse instead. BTW, how much disk space does the IE9 image need? Anomie⚔ 00:50, 5 September 2012 (UTC)- I haven't succeeded in getting qemu to work with KVM, and without KVM the virtual PC images aren't going to work for me. Thanks anyway. The IE9 image takes up about 13.7 GB. —Remember the dot (talk) 21:35, 8 September 2012 (UTC)
- All I can tell you about KVM is to make sure you have the proper kernel support compiled in or loaded as a module (you should have /dev/kvm), and a CPU with AMD-V or Intel VT-x (check /proc/cpuinfo for "svm" or "vmx"). Anomie⚔ 07:39, 9 September 2012 (UTC)
- Well, that explains the mystery. My processor has neither svm nor vmx. —Remember the dot (talk) 19:01, 9 September 2012 (UTC)
- All I can tell you about KVM is to make sure you have the proper kernel support compiled in or loaded as a module (you should have /dev/kvm), and a CPU with AMD-V or Intel VT-x (check /proc/cpuinfo for "svm" or "vmx"). Anomie⚔ 07:39, 9 September 2012 (UTC)
- I haven't succeeded in getting qemu to work with KVM, and without KVM the virtual PC images aren't going to work for me. Thanks anyway. The IE9 image takes up about 13.7 GB. —Remember the dot (talk) 21:35, 8 September 2012 (UTC)
- I've been using the IE8 image for occasional testing for some time now with qemu (specifically,
- I tried the IE8 and IE9 images with VirtualBox but both gave me a "corrupted ACL" error and wouldn't boot. Have you had better luck? —Remember the dot (talk) 23:12, 4 September 2012 (UTC)
- They've got a bloated IE9 image listed there, no IE10 yet. Anomie⚔ 01:53, 3 September 2012 (UTC)
You could set up a partition on your hard drive for Windows 7. It works much better than Win XP as concerns not screwing up other O.S. partitions. I have a Win XP partition and a Win 7 partition on my hard drive. I have not used the Win XP partition though for a long time. But Windows 7 takes care of all the multi-booting choices and interfaces. I don't know though how Win 7 and Linux partitions play together though. But I read that it is common for people to have both. --Timeshifter (talk) 04:30, 3 September 2012 (UTC)
- That means you have to buy (or steal) Windows 7. And shut everything down every time you want to test something in Windows, or set up something like Xen. Anomie⚔ 10:55, 3 September 2012 (UTC)
Numeric entities
First of all, thanks for your great syntax highlighter script!
I noticed that only named entities are highlighted, but not numeric: no break - no break - no break
The end of the long regular expression should be something like &(?:\\w+|#\\d+|#[xX][0-9a-fA-F]+);
instead of just &\\w+;
. --Schnark (talk) 08:17, 6 September 2012 (UTC)
Unresponsive script
To reproduce this bug in your syntax highlight script, remove the pre tags from the following code:
{| style="border: 1px dashed green;" |- | ohne <code>cellpadding</code> und ohne <co |} {| style="border: 1px dashed green; padding: 10px;" |- | <code>padding: 10px;</code> |}
I was adding something in the first table. After I had opened the new code tag, my Firefox 10 stopped in the line
if (match[0] == stopAfterText)
and asked me, whether I want to stop the unresponsive script or not. Even stopping didn't help, but finally at least I managed to type a closing >, and everything worked again. --Schnark (talk) 09:41, 8 September 2012 (UTC)
- Thank you for the two bug reports! This is exactly the kind of feedback I need. Both should be fixed now. Also, I highly recommend updating to the latest version of Firefox for maximum JavaScript performance. —Remember the dot (talk) 16:23, 8 September 2012 (UTC)
RfC on differentiating reference syntax in text window
Hi Remember the dot-- based on the village pump discussion on giving reference syntax a unique color to differentiate from other text while editing, I've opened up an RfC to expand the audience on the topic. You are welcome to participate anytime. I also want to say your tool has been a pleasure to use, and I hope it can be expanded to other editors. I think it can be a real game changer. I, Jethrobot drop me a line (note: not a bot!) 00:26, 21 September 2012 (UTC)
Some suggestions to improve your syntax highlighter
In
else //|
{
//table
writeText("{{", syntaxHighlighterConfig.tableColor || color);
the last line should be
writeText("{|", syntaxHighlighterConfig.tableColor || color);
In
if (tagName == "nowiki" || tagName == "pre" || tagName == "syntaxhighlight")
you should add source
, math
, hiero
, and perhaps timeline
(wikilinks are allowed, but everything else is a big mess).
In
stopAfter.replace(/([\|\[])/g, "\\$1")
you should escape more, I once found a JavaScript error in the log about an invalid regular expression. Try <**. Since it is called inside a setIntervall
this error didn't break anything, but should be fixed anyway. You could use
$.escapeRE(stopAfter)
Another solution would be to limit the tag name to \w
, this also makes raw less than signs possible. You could use <(?:\w+|!--[\\s\\S]*?-->)
as regular expression to search for tags, and set tagName
to what this expression found, and just look for attributes and selfclosing tags from that point on.
To allow <br>
you could simply add
if (tagName == "br" || tagName == "hr") emptyTag = true;
For some reason no scroll event seems to be fired when the textbox scrolls after you click a button in the toolbar, so scrollTop
should be synced regularly.
I don't mind the font size, but perhaps 10pt
is a bit too small for some people. Is there any reason, why you set both wpTextbox1
and wpTextbox0
to the same fixed value, instead of just copying the style of wpTextbox1
to wpTextbox0
?
I know that caching isn't easy, but I think that it should be possible for discussions (which are at least for me pages where I regularly reach the 150ms limit):
- Save the original value of the text and the CSS for its syntax highlight.
- When highlight the text next time, first test whether the text still starts with the original text, and just append the CSS for the added text to the original CSS instead of parsing the whole text again.
This might cause some problems when there is some unclosed structure in the original text, but this should happen very rarely, and when the user inserts the missing closing tag before he adds his own comment, the script will be forced to parse the complete text anyway. Perhaps there even is a good way to test for such unclosed things. --Schnark (talk) 08:02, 24 September 2012 (UTC)
- Hi again Schnark, thanks for the help and thanks for taking the time to dig through some pretty confusing code. I fixed the issues you found with tables, tags that should be treated like
<nowiki>
, custom font sizes (though I think the default font size is too small), and escaping the regular expression. Because escapeRE is a one-line function, I just copied its code to make off-wiki testing easier.
- I don't really want to add a special case for
<br>
and<hr>
; I find these syntaxes confusing and would prefer to encourage the terminated forms<br/>
and<hr/>
.
- I could not reproduce the bug you found with the editing toolbar. Does it happen only in certain circumstances?
- Finally, I experimented with caching the text that has already been parsed but did not come up with a good implementation. One day we may want to rewrite the parser altogether. So, improvements in this area are probably not going to happen in the near future, but I recognize that it is a sore spot. Thanks again for the help! —Remember the dot (talk) 23:56, 25 September 2012 (UTC)
- Since I use a script that automatically replaces all <br> with <br /> I actually don't care about whether it is highlighted or not.
- I currently can't test the script since for some reason the only browser installed on this computer are IE9 and FF3. I think the script could work in Firefox 3 if you add
false
as the third argument to alladdEventListener
, see [1], andnull
as second parameter togetComputedStyle
([2]). But if you don't want to care about old browsers just ignore them, I try to find a computer with reasonable browsers. - If I remember correctly, the scrolling bug was easily reproducible in FF 15:
- Click edit for this section, scroll the editbox such that the last line is at the very bottom.
- Fill this last line almost completely.
- Click "Signature and timestamp"
- The signature doesn't fit into the last line, the editbox scrolls one line, but the syntax highlight stays where it was. — Preceding unsigned comment added by Schnark (talk • contribs) 09:35, 26 September 2012 (UTC)
Another example of syntax highlighting on an online service
There's an interesting text highlighter (and you might be aware of it) at this website called jsFiddle, which highlights syntax in HTML, javascript, and CSS. Try it out, if you haven't-- it seems to work very well. They also have documentation here. Let me know what you think. I, Jethrobot drop me a line (note: not a bot!) 05:32, 5 October 2012 (UTC)
- Hey, thanks for the link. They definitely have a different approach than me, though I can't immediately tell how it works. By the way, I have been watching the RfC on syntax highlighting. I have not participated because I feel that the discussion is premature without a working implementation, and I am still hoping that the browser compatibility issues holding up my highlighter will be resolved. That said, it has really been interesting to see these other possible implementations come out of the woodwork, and I look forward to seeing where the discussion goes. —Remember the dot (talk) 05:12, 9 October 2012 (UTC)
- Well, I certainly hope you can make yourself a part of the effort in making the implementation, whenever we get to that stage. I think the syntax highlighter is fantastic, and I hope it makes its way, in any form, into more widespread use here. Is there anything you can add to the discussion in terms of a proposed implementation? That seems to be what most editors are concerned about, and understandably so. I, Jethrobot drop me a line (note: not a bot!) 05:57, 9 October 2012 (UTC)
- On no: there is a gadget using a similar framework for syntax highlighting, AceWikiEditor. It's fast, but the highlighting has several bugs, especially it can't handle nested tables etc. --Schnark (talk) 09:26, 10 October 2012 (UTC)
Dropdowns from edit toolbar hide behind higlighted text
As the title says: Open the "Advanced" or "Cite" toolbar from the default edit toolbar and click on the dropdown. It will hide behind the edit box. This sometimes happens (happened?) even without your script, but you should add something like $('.tool-select *').css({zIndex: 5});
in the initialising code. --Schnark (talk) 09:38, 6 October 2012 (UTC)
PS: I'm doing some experiments in User:Schnark/vector.js. It still has too many bugs to be usefull, but sometimes I'm able to edit Barack Obama with live syntaxhighlight in ~ 70ms.
Some notes about performance
My version of your syntax highlighter (still at User:Schnark/vector.js) now seems to be fast enough to be useful, editing the complete Barack Obama works well. If you want to copy parts (or the whole) of it, feel free to do so!
Two points about performance you should note anyway:
- I still can reproduce the scrolling bug I mentioned above. But the only solution is to sync the
scrollTop
every time the text is highlighted. This is terribly slow, so I decided to ignore the bug. - The first run of the highlighter might take more than 150ms on long articles. Your script should ignore this: One reason for the long time is the creation of all the
<span>
s. Once they are created the script is faster, so you should only abort when the script takes too long for the second time. --Schnark (talk) 08:26, 8 October 2012 (UTC)
- Hey Schnark, thanks for all the work you've put into this. Did you write that entire parser yourself? It is impressive, and it would make sense that it is faster than my regex-based one. If your version wins out, I wouldn't mind. You certainly seem to have a lot more energy to devote to this than I do...
- I did fix the scroll synchronization issue by polling every 500ms for changes to scrollTop. This is similar to how the highlighter script detects that another script has edited something without generating an input event. I thought it was a reasonable compromise between continual polling and no polling. —Remember the dot (talk) 05:12, 9 October 2012 (UTC)
- I must admit that my parser (now at de:Benutzer:Schnark/js/syntaxhighlight.js) looks faster than it actually is. The first version took twice as long as yours. Since it is quite easy to add things to my parser, I'm tempted to do so, more stuff takes more time, so I'm not sure whether your script is actually faster. Since I splitted the code into different parts, I'm able to see the time each part takes individually: The two most significant operations seem to be creating the
<span>
s that get highlighted and creating and applying the CSS. I'm going to make some more tests to see how to make things faster, but I'm not sure whether this is really possible. --Schnark (talk) 09:20, 10 October 2012 (UTC)
- I must admit that my parser (now at de:Benutzer:Schnark/js/syntaxhighlight.js) looks faster than it actually is. The first version took twice as long as yours. Since it is quite easy to add things to my parser, I'm tempted to do so, more stuff takes more time, so I'm not sure whether your script is actually faster. Since I splitted the code into different parts, I'm able to see the time each part takes individually: The two most significant operations seem to be creating the
Orphaned non-free media (File:Koch Industries logo.png)
Thanks for uploading File:Koch Industries logo.png. The media description page currently specifies that it is non-free and may only be used on Wikipedia under a claim of fair use. However, it is currently orphaned, meaning that it is not used in any articles on Wikipedia. If the media was previously in an article, please go to the article and see why it was removed. You may add it back if you think that that will be useful. However, please note that media for which a replacement could be created are not acceptable for use on Wikipedia (see our policy for non-free media).
If you have uploaded other unlicensed media, please check whether they're used in any articles or not. You can find a list of 'file' pages you have edited by clicking on the "my contributions" link (it is located at the very top of any Wikipedia page when you are logged in), and then selecting "File" from the dropdown box. Note that all non-free media not used in any articles will be deleted after seven days, as described on criteria for speedy deletion. Thank you. Hazard-Bot (talk) 04:09, 11 October 2012 (UTC)
Talkback
Message added 05:38, 8 November 2012 (UTC). You can remove this notice at any time by removing the {{Talkback}} or {{Tb}} template.
I, Jethrobot drop me a line (note: not a bot!) 05:38, 8 November 2012 (UTC)
dragging text
Hi there,
it seems dragging selected text in the textbox doesn't work well if the highlighter is at work (even if both the selection and target areas aren't highlighted). It works otherwise if the text contains no mediawiki syntax. Issue is occurring in Firefox 16 on Ubuntu 12.04, using the modern skin with some modifications. Do you think it could be fixed? --Waldir talk 15:32, 9 November 2012 (UTC)
- It works perfectly for me, Firefox 16 on Ubuntu 12.10. I'd need more details about what you're seeing to know what the problem might be. My guess is that one of your other scripts is interfering with the syntax highlighter--you'll have to try removing each one individually to see which one is causing the problem. —Remember the dot (talk) 18:46, 10 November 2012 (UTC)
- That's strange, I'm testing this again right now and it doesn't occur anymore. Anyway, sorry for the false alarm. If I see the issue again I'll try to do some more thorough testing and will let you know if I find anything relevant for the script. Cheers, Waldir talk 23:26, 11 November 2012 (UTC)
Notice of change
Hello. You are receiving this message because of a recent change to the administrator policy that alters what you were told at the time of your desysopping. The effect of the change is that if you are inactive for a continuous three year period, you will be unable to request return of the administrative user right. This includes inactive time prior to your desysopping if you were desysopped for inactivity and inactive time prior to the change in policy. Inactivity is defined as the absence of edits or logged actions. Until such time as you have been inactive for three years, you may request return of the tools at the bureaucrats' noticeboard. After you have been inactive for three years, you may seek return of the tools only through WP:RFA. Thank you. MBisanz talk 00:20, 4 December 2012 (UTC)
Syntaxhighlighter- duplicate edit window
In case this comes up: If Syntaxhighlighter is enabled as a Gadget and in your JS, then the edit windows gets duplicated with the highlighting but no text. --— Gadget850 (Ed) talk 12:47, 6 December 2012 (UTC)
- Good to know, thanks. —Remember the dot (talk) 05:44, 7 December 2012 (UTC)
Featuring syntax highlighter in a Signpost series.
I'm working on a Signpost series to bring attention to new wiki-tools and feedback to their authors. So far, I have Nettrom, author of SuggestBot, West.andrew.g, author of WP:STiki, and EpochFail (myself), author of WP:Snuggle. I'm hoping to convince you to include an article on your syntax highlighter. The idea is still pretty preliminary, but I've started working on a draft article for Snuggle that you can look at to see what I have in mind. --EpochFail(talk|work) 15:42, 13 December 2012 (UTC)
- I'd be happy to! Here is my draft for you. Feel free to make improvements and don't worry about copying over my edit history when you put it into the Signpost. —Remember the dot (talk) 21:24, 16 December 2012 (UTC)
markup precedence?
Hi, I noticed that some forms of markup seem to take precedence over others, resulting in things like this. Maybe that's inevitable, but I thought you might want to take a look. --Waldir talk 20:14, 17 December 2012 (UTC)
- Hi Waldir, I'm glad you've been enjoying the highlighter. That is an interesting problem you ran into. To make the syntax highlighter work at a reasonable speed, its parsing algorithm is very simple. Basically what it sees is this:
<tag><template>content</tag><tag>content</template></tag>
- And because the syntaxes are not symmetric, it considers them to be malformed just like if they were asymmetric HTML tags. The real problem here is that {{expand language}} was using needlessly convoluted syntax. I have since corrected it. Now the syntax highlighter sees:
<tag><template>content</template></tag>
- Let me know if you find any other templates like this, and thanks for the feedback! —Remember the dot (talk) 04:36, 18 December 2012 (UTC)
- Awesome. Indeed I always thought some of the subst constructs I see around looked awfully convoluted, but I never had the guts to change them just in case they depended on some obscure software feature that would break unless used that way.
- On a totally unrelated note: I see that '''''' is "self-closing". Is there a way '''' could be too? Does that make sense given the logic the highlighter is using? --Waldir talk 06:24, 18 December 2012 (UTC)
- Ah, that's another interesting case. In reality both '''''' and '''' turn into a single quotation mark ' if you try to use them. So really, because they are not self-closing in MediaWiki, neither should be self-closing in the highlighter either. That said, for performance and to discourage sloppy syntax, the syntax highlighter's behavior in these cases is undefined. —Remember the dot (talk) 01:35, 19 December 2012 (UTC)
- I see. Makes sense. It's pretty much an edge case anyway. Thanks for the clarification. --Waldir talk 05:42, 19 December 2012 (UTC)
FYI, Here's another one that confuses the highlighter. Btw, happy new year! --Waldir talk 15:49, 1 January 2013 (UTC)
- Thanks for fixing that one too. I was unsure about the (undocumented, as you noted in the template's talk page) subst parameter. Cheers, and happy new year! --Waldir talk 14:54, 4 January 2013 (UTC)
Code in maintenance templates
I see you removed </includeonly><includeonly>
from many maintenance templates, like in this edit. These were made by many very experienced template editors. Are you sure this was a good idea? Debresser (talk) 16:22, 19 December 2012 (UTC)
- (talk page stalker) it looks to me as if those very experienced template editors may have made syntax errors. The edits have removed the turning off and them immediate turning on of "includeonly" and this edit should have no effect.
subst = <includeonly>{{subst:</includeonly><includeonly>substcheck}}</includeonly>
- has the same effect as
subst = <includeonly>{{subst:substcheck}}</includeonly>
- There is, of course, the possibility that I am mistaken, and I will be content to be educated if so. Fiddle Faddle (talk) 17:10, 19 December 2012 (UTC)
- I understood the logic myself, and wondered about this myself in the past. But I never checked it. I ran a test, and it seems that it is okay. So let's hope so. Debresser (talk) 17:33, 19 December 2012 (UTC)
- I'm glad you ran a test. I became experienced in redundant tags when converting documents from M$ Word to HTML, so can spot this stuff at 1,000 paces :) Fiddle Faddle (talk) 17:35, 19 December 2012 (UTC)
- Thanks for chiming in, Fiddle Faddle. I did test the behavior before making the edit, and after that I started by changing just one relatively high-use template and waiting a day to see if anything broke that I didn't notice. Still, I'm sure there are thousands of other templates with this or other mistakes. This particular mistake was caught by the syntax highlighter gadget (see the above thread...) —Remember the dot (talk) 01:46, 20 December 2012 (UTC)
- Well, we are in the world of Mediawiki, where ridiculous things happen because the underlying engine is unusual at best! Logically nothing can break when making this change. I suggest you unleash the beast and quietly make this correction wherever you feel appropriate. I was simply amused at the 'many very experienced editors' concept. This is Wikipedia! We have many editors with great experience, and a good few with great expertise. The intersection of great experience experience and great expertise is not a large intersection :) Fiddle Faddle (talk) 10:39, 20 December 2012 (UTC)
Neat script
Great job. I love the syntax highlighter. It makes seeing certain mistakes so easy that I may never do them again (e.g., unpaired tags are now so obvious). So here's a baby sugar glider to show my appreciation.--Fuhghettaboutit (talk) 00:26, 28 December 2012 (UTC) |
- I'm glad you like it! I think that avoiding mistakes like unpaired tags will make the code a lot more readable for newcomers. Thanks for the baby sugar glider, and happy new year! —Remember the dot (talk) 03:37, 29 December 2012 (UTC)
- Hey again. In making a recent edit I noticed an issue. I often use <p> when I write talk page posts to break up thoughts or large posts – especially when I am indenting and want a line break, which mean there's no need to hit return twice and insert another colon. Anyway, I noticed that when you use this code it treats it as if it's the type of code that is going to have an ending </p>, highlighting all of the subsequent contiguous content and thus obscuring subsequent code highlighting. So I would suggest that it treat <p> as a self-contained coding instance, highlighting just that code and not subsequent text. Best regards--Fuhghettaboutit (talk) 05:17, 4 January 2013 (UTC)
- I'm glad you like the script! It has been quite the challenge to get it to highlight without introducing noticeable delay, and one of the ways it accomplishes this is through a very fast and simple parsing algorithm. Were I to start carving out exceptions, the performance of the script would degrade significantly. More importantly though, it's just good practice to press Enter and indent again rather than throwing a
<p>
into the middle of a block of text, as it makes the wikitext more understandable when others open it for editing. For example, what if someone wanted to insert a reply to the first paragraph before the second paragraph, as Wikipedians often do? —Remember the dot (talk) 05:43, 4 January 2013 (UTC)- I do indeed! If feasibility is an issue on this, then so be it. But I think you may have misunderstood: I was speaking specifically, as in the linked example of where this occurred, about usage in talk page posts. I would never use this (nor should it be used) in article content, which your criticisms above would seem to speak to. If anyone tried to interlineate a post in between two paragraphs of a single post of mine that ended in my signature (which I can't remember the last time I saw someone do, to me or to anyone else) then I'd probably whack them with a wet trout. And I'm not sure why anyone would need to look at the wikitext of a user's talk page post for understanding. Cheers!--Fuhghettaboutit (talk) 05:58, 4 January 2013 (UTC)
- The wikitext's readability does matter even on talk pages, especially for newcomers who undoubtedly find wikitext strange and complicated. Using a consistent style in both articles and everywhere else will help. There are also people who like to glance at the post above theirs in the text editor while composing a reply. If after taking a good hard look at what you're doing you really need to include a
<p>
, just make sure to end it properly with</p>
. As you pointed out, the highlighter helpfully discourages stray<p>
tags in articles, and I don't want to change that by making an exception. - By the way, I have more than once had a discussion broken into threads by paragraph, though perhaps this is a fading custom. —Remember the dot (talk) 06:20, 4 January 2013 (UTC)
- The wikitext's readability does matter even on talk pages, especially for newcomers who undoubtedly find wikitext strange and complicated. Using a consistent style in both articles and everywhere else will help. There are also people who like to glance at the post above theirs in the text editor while composing a reply. If after taking a good hard look at what you're doing you really need to include a
- I do indeed! If feasibility is an issue on this, then so be it. But I think you may have misunderstood: I was speaking specifically, as in the linked example of where this occurred, about usage in talk page posts. I would never use this (nor should it be used) in article content, which your criticisms above would seem to speak to. If anyone tried to interlineate a post in between two paragraphs of a single post of mine that ended in my signature (which I can't remember the last time I saw someone do, to me or to anyone else) then I'd probably whack them with a wet trout. And I'm not sure why anyone would need to look at the wikitext of a user's talk page post for understanding. Cheers!--Fuhghettaboutit (talk) 05:58, 4 January 2013 (UTC)
- I'm glad you like the script! It has been quite the challenge to get it to highlight without introducing noticeable delay, and one of the ways it accomplishes this is through a very fast and simple parsing algorithm. Were I to start carving out exceptions, the performance of the script would degrade significantly. More importantly though, it's just good practice to press Enter and indent again rather than throwing a
- Hey again. In making a recent edit I noticed an issue. I often use <p> when I write talk page posts to break up thoughts or large posts – especially when I am indenting and want a line break, which mean there's no need to hit return twice and insert another colon. Anyway, I noticed that when you use this code it treats it as if it's the type of code that is going to have an ending </p>, highlighting all of the subsequent contiguous content and thus obscuring subsequent code highlighting. So I would suggest that it treat <p> as a self-contained coding instance, highlighting just that code and not subsequent text. Best regards--Fuhghettaboutit (talk) 05:17, 4 January 2013 (UTC)