Forums Feature Suggestions Missing Documentation

$window().cx $window().cy, $mouse.cx and $mouse.cy are missing from the help file.

7.52's /fupdate and 7.53's $caller not documented in /help

also, if the /fupdate setting isn't going to be saved to mirc.ini [options], could there be something like an $fupdate identifier which returns the 0-100 value shown by /fupdate? This would allow scripts to save/restore the existing value, so they could change it during a specific burst of output.

/help /dll says:



Could the dll page be updated from '900 chars' to indicate the current max limit in bytes, now that Unicode means chars and bytes aren't the same thing?

Most DLL's won't care as long as the length is longer than 8292, but the info would be useful for a DLL which needs to support languages from the unicode codepoints above 2047 where each 'char' requires 3 bytes.

I've only tested 'data but not 'parms', but 900 appears to be an obsolete number from older 6.x versions which had that shorter line length. I tested to see how long of a text string of hex digits could be written into 'data' for sending back to mIRC, before it crashes mIRC. The byte length in 6.35 seems to be a few dozen bytes longer than than the 4150 limit. In 7.52 the byte length limit was somewhere around 8400 bytes, I'm assuming this extra length is for supporting multi-byte unicode characters. And now that the line length limit has doubled, the valid string bytes-length has doubled above 16k.

In the mirc 7.56 there's no help info about $urlget. It will be good if the mirc.chm will be updated in time. I don't know about other changes/adds if they are

/help $ltimer

both locations says it returns the number of the timer. Should say 'timer ID'

//timerfoobar $+ $rand(1,99) 1 1 noop | echo -a $ltimer

$file().attr is returning a bunch of letter which are different from what one can read on the internet.
I believe mIRC is translating the constant here https://docs.microsoft.com/fr-fr/windows/win32/fileio/file-attribute-constants to various letters, if some can be obvious, others aren't, it would be great to have a documentation of each possible letter.

It looks like $urlget can return 0 when the connection is refused or something. I'm doing a lot of successive $urlget call via milliseconds timers and at some points i get 0 from $urlget, I'm not sure if this is a bug but I'm assuming the website is refusing my connecting as an antispam protection and that cause $urlget to return 0. My point is that in this case the callback alias won't be triggered, all of this is not documented, it would be nice to know what 0 really means and how it's not going to trigger the callback alias.

/loadbuf -g

-g = ?? (it's undocumented)

Missing $evalnext() identifier

Examples should be changed since root folder of c:\ has been read-only for a long time, unless mIRC is running in $isadmin=$true mode. It's hard to tell if there are other similar path strings elsewhere in /help since the search function returns a lot of false matches

/write -c c:\info.txt This file will be erased and have this line written to it
/write -c c:\info.txt This file will be erased and have this line written to it
/write -il5 c:\info.txt This line will be inserted at the 5th line in the file
/write -dl5 c:\info.txt
/write -dstest c:\info.txt

edit: this is in response to someone coming into channel asking why the /write command in the /help doesn't work

Thanks, I've removed all references to c:\ from the help file.

While making this answer to a question
https://forums.mirc.com/ubbthreads....-you-must-specify-a-parameter#Post268998

I noticed that labels in popups aren't behaving the same way as the command following the semi-colon.

When more than 1 nick is highlighted, the command following the semi-colon sees $2 $3 etc get filled in, but they are not filled as seen by the menu label. While it is true that the /help doesn't mention anything beyond $1, the default popups menu does contain them and the command here does use them:

.Op:/mode # +ooo $$1 $2 $3

... and users could see unexpected behavior if they assume that $2 behaves the same way across the entire label if you have the following lines in the nicklist popup menu:

Label (0) $0 (1-) $1- parms $parms (2) $2 : echo -a (0) $0 (1-) $1- parms $parms (2) $2
.Give Op to $1-3:/mode # +ooo $$1 $2 $3

In the label, the $0 is always 1 when $2- is filled due to more than 1 nick selected, $parms is always ignoring the presence of $2- nicks, and $2- $2 $3 etc are always blank. This can result in a menu command targeting a different group of nicks than indicated by the label, ie giving ops to 3 nicks even though the label says only 1 nick.

Whether $2- should become part of the menu label, or the /help should be updated to indicate that $2- are valid in the :/command portion, I dunno.

$lock(sendmessage) ; 7.34
$min()/$max() .text .textcs ; 7.62
/hotlink -dm ; 7.23

It appears that $lock is having backwards compatibility for obsolete functionality, where it returns $false for things no longer able to be disabled from the menu. This $false reply also happens for things including:

//echo -a $lock(decode) $lock(run) $lock(com) $lock(dll)

My reply wasn't completely accurate. $lock(sendmessage) is not one of the other obsolete features returning $false. Instead, the option to enable or disable it is on options/other, instead of being at options/other/lock along with the other active remaining items.

Based on the different behavior of these 2 commands, it appears that 'v' is an undocumented switch, so it would be great if F1 help could tell us what -v does.

/echo -jkopuwxyz
/echo -jkopuvwxyz

If -v turns out to be a 'do nothing switch letter that will never be used as a valid switch' as referenced HERE that would also be helpful if it were documented.

The documentation for the /*line commands could be made more clear by having the syntax line showing the location for the switch letters. Especially since it's different than for /echo because the switch letter goes before the color index instead of after.

i.e.
/aline -a 4 abc
vs
/echo 4 -a abc

It's probably from /drawtext -v &binvar being passed to /echo internally. Or it's an experiment where Khaled tried, or anticipated, adding &binvar support to /echo in order to print things like sequential spaces.

$read documentation vs behavior

I assume at this point the fix is to clarify F1 help rather than to change the behavior of $read, since that would behavior that's not backwards compatible.

The description of $read's 't' switch makes users understand it to be saying that that avoiding using the 't' switch while the top line of a text file is an integer - would allow the 's' and 'w' and 'r' switches to mimic the behavior of $read(file,n) where a small integer as the top line of a text file could allow the search range to be restricted across only a portion of the file.

Instead, using the 's' switch with/without the 't' switch has identical behavior except for having $readn returning a decremented integer if the 't' switch isn't used when the top line is an integer.

i.e. I was expecting the next example to fail to find a match since the match wasn't within the 1st 2 lines, but instead the only difference is whether $readn returns 5 or 4:

//write -c foobar.txt $+(2,$crlf,a1,$crlf,b2,$crlf,c3,$crlf,xyz aaa) | echo -a $read(foobar.txt,nts,xyz) $readn : $read(foobar.txt,ns,xyz) $readn

I don't know whether this would deserve having the t switch changed to be a tN the same way there's a wN subswitch.

The beta changelog does not mention two new props for $notify, $notify().away $notify().awaymsg.