mIRC Home    About    Download    Register    News    Help

Print Thread
Page 1 of 2 1 2
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Yes, I know that mIRC now has full unicode support. But the fact does not explain why it cannot display national characters when started under cp1251 locale. 6.35 works just fine.

Example screens:

wrong (7.x)
correct (6.35)

Many so-called "full unicode apps" work fine with cp1251 here, why mIRC can't? The problem is that many IRC-server here, in russia, has issues with implementing UTF-8 suppport on IRCd-side, so far only few small networks support it in some way.

Many conditions makes migration impossible or too complicated.

Info:
beta27 7.04 Vista 0 0 1

Related reports:
http://forums.mirc.com/ubbthreads.php?ubb=showflat&Number=221956&page=1#Post221956


"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Oct 2003
Posts: 3,918
A
Hoopy frood
Offline
Hoopy frood
A
Joined: Oct 2003
Posts: 3,918
This issue has already been raised, discussed, and commented on by Khaled. A duplicate report won't change much; Khaled is aware of the compatibility issues. Also note that the comments in that thread are relevant to what your options are. There will be ways to support your legacy encodings "soon".

It should also be noted that you specifically set your encoding to one that mIRC no longer supports ("CP1251 is your translation scheme"). Perhaps you should consider switching the encoding on the server end to UTF-8. You claim it's not "properly implemented", but I don't know of any specific issues.

As far as "many conditions make migration impossible"-- if your definition of "migration" is continuing to use CP1251, well, we have different definitions of "migration". The only real migration, if what you said about server-side UTF-8 is true, is to get your servers in Russia to properly support Unicode. Until that happens, "migration is impossible"-- but mIRC is not the one holding you back. it's been almost 14 years since Unicode was standardized; it's been by and large the standard on the web for the last ~6. IMHO there's really no reason for mIRC to wait any longer for those servers to get their act together.


- argv[0] on EFnet #mIRC
- "Life is a pointer to an integer without a cast"
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Quote:
This issue has already been raised, discussed, and commented on by Khaled. A duplicate report won't change much; Khaled is aware of the compatibility issues. Also note that the comments in that thread are relevant to what your options are. There will be ways to support your legacy encodings "soon".

Where? Linkify me please? Wasn't able to find it.
UPD: nvm, found it.

Quote:
Perhaps you should consider switching the encoding on the server end to UTF-8. You claim it's not "properly implemented", but I don't know of any specific issues.

80% of IRC-servers (networks) I do use has no support for UTF-8, other 20% has experimental support and those servers/networks aren't so important for me.
I hope now you can see the size of issue :<

Quote:
As far as "many conditions make migration impossible"-- if your definition of "migration" is continuing to use CP1251, well, we have different definitions of "migration".

I was talking about old mIRC -> new mIRC migration.

Quote:
The only real migration, if what you said about server-side UTF-8 is true, is to get your servers in Russia to properly support Unicode.

We're talking not only about the russian part of IRC world. How many (in %) other national servers support UTF-8? Greek, Italian, German, etc? Btw, UTF-8 support was initially (firstly) implemented in russian edition of well-known IRCds because of obvious reasons. Popularized by UnrealIRCD bynets edition.

Quote:
Until that happens, "migration is impossible"-- but mIRC is not the one holding you back.

Huh? I never said mIRC holding something back. Perhaps there is some kind of misunderstanding? See my reply about migration above.

Quote:
And, IMHO, it's been almost 14 years since Unicode was standardized; it's been by and large the standard on the web for the last ~6. There's really no reason for mIRC to wait any longer for those servers to get their act together.

It is much easier to add backward-compatibility into most popular IRC-client instead of asking network administration of every IRC-network to implement UTF-8 support and wait 'till it will be done.
Don't you agree? >_>

Last edited by Lusht; 05/06/10 09:59 PM.

"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Jul 2006
Posts: 248
B
Fjord artisan
Offline
Fjord artisan
B
Joined: Jul 2006
Posts: 248
If you use ZNC 0.090 (rc1 is out currently), you can do

/znc loadmod charset -force UTF-8 CP1251

which will enable you to use mIRC in UTF-8 mode from ZNC to mIRC, and the ZNC <-> IRC server connection will use CP1251 transparently.

Joined: Oct 2003
Posts: 3,918
A
Hoopy frood
Offline
Hoopy frood
A
Joined: Oct 2003
Posts: 3,918
Originally Posted By: Lusht
It is much easier to add backward-compatibility into most popular IRC-client instead of asking network administration of every IRC-network to implement UTF-8 support and wait 'till it will be done.
Don't you agree? >_>


Quite simply: no, I don't agree.

mIRC had support for codepages at the loss of proper Unicode support. mIRC decided to do one thing well and now supports Unicode properly at the loss of codepages. In theory, backwards compatibility is nice-- in practice, it's one or the other. This is likely why your IRC servers have "experimental" UTF-8 support; because it's complicated to support multiple encodings. It's actually easier on the server end, but that's a discussion for ircd implementors, not this forum.

If mIRC doesn't ask networks to start supporting UTF-8, it will never get done. It's not considered "backwards-compatibility" if mIRC plans on supporting codepages forever; so if mIRC doesn't start pushing a little harder, the networks won't do anything. It's important to point out that even mIRC has supported UTF-8 for almost 4 years now. The Unicodization of mIRC was impending for many years now, and your networks did absolutely nothing-- that's embarrassing. Clearly mIRC needs to start taking a more aggressive stance, as the more popular IRC client out there.

mIRC wants to do the right thing and move people away from codepages and onto a proper encoding; the encoding that every modern OS supports, almost all of the web uses, and most other IRC clients use. I'm talking about Unicode, of course. The percentage of support on the servers you use might be low right now, but by releasing mIRC 7, this will change. That's a good thing.

As a sidenote, mIRC is not forcing you to immediately upgrade to 7.1 as soon as it is available. 7.x has been put out there as a beta to show people how mIRC will be working. This gives your network admins time to start improving their Unicode support. They may not be done by 7.1, but you can continue to use 6.35 and tell them to start taking Unicode more seriously.

You shouldn't be complaining about the lack of support for codepages here-- you should be complaining about the lack of support for UTF-8 on your network. If you took the time to write them the same kind of posts you wrote here, maybe they would actually do something about it, rather than having you wait indefinitely.


- argv[0] on EFnet #mIRC
- "Life is a pointer to an integer without a cast"
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Quote:
Quite simply: no, I don't agree.

Ok, you don't see a problem here. but I do.
...
What now? If you don't see a problem - it (a problem) does not exist then?

Quote:
This is likely why your IRC servers have "experimental" UTF-8 support; because it's complicated to support multiple encodings.

No, because UTF-8 implementation is not perfect and still in discussion. There is no "good" way to do it, no one perfect mechanism. That's why on some servers it's implemented in different ways.
And no, "mine" IRC-servers support many codepages because of backward-compatibility I was talking about....

Quote:
It's actually easier on the server end, but that's a discussion for ircd implementors, not this forum.

it is fun actually, how different kind of people see the solution for one problem. IRCd-devs, for example, has different opinon:

(03:12:11) <+Kein> Brain Will be InspIRCD support iconv by default? It's is a codepage encoding and support for some codepages, like UTF-8, cp1251 (russian), KOIR-8 and more on iRCD-level...
(03:12:11) <+Kein> Like in the ByNets Unreal distro
(03:12:50) <+Adremelech|Laptop> that sounds like a crossbreed between "ick" and "eewww"
(03:13:13) <+Adremelech|Laptop> "eeecccckkk"
(03:13:15) <+Adremelech|Laptop> laugh
(03:16:34) * A (cash082@ChatSpike-c0b9b815.fios.verizon.net) has joined #inspircd
(03:16:35) * ChanServ sets mode: +v A
(03:17:18) <~Brain> lo A
(03:17:29) <~Brain> Kein: no
(03:17:49) <~Brain> a client should do codepage translation, rfc says only one specific charset, finnish one i think
(03:18:58) <+Kein> But they shouldn't.
(03:19:04) * +typobox43 (typobox43@nin10doh.com) Quit (Ping timeout: 121 seconds)
(03:20:40) <~Brain> yes they should
(03:20:48) <~Brain> an ircd just relays data
(03:20:50) <~Brain> it doesnt translate it
(03:21:14) <+Kein> Sorry, i mean: but they don't do it.
(03:21:16) <~Brain> the protocol itself isnt compatible with a whole load of character sets
(03:21:20) <~Brain> they should
(03:21:24) <~Brain> not our job
(03:21:25) <~Brain> :p
(03:21:41) <~Brain> nor do i have any desire to touch any bloated nonportable library such as iconv
(03:21:53) <~Brain> along with all the shitty deps it needs


And ircclient-devs (well, lets pretend for a moment you have the right to speak for Khaled/mIRC dev team) thinks it must be IRCD job.
DEAD END. Infnite loop. Problem exist, but no one want to fix it ;<

Quote:
mIRC wants to do the right thing and move people away from codepages and onto a proper encoding; the encoding that every modern OS supports, almost all of the web uses, and most other IRC clients use. I'm talking about Unicode, of course. The percentage of support on the servers you use might be low right now, but by releasing mIRC 7, this will change. That's a good thing.

That's a good thing, right. But let me show an anology for such situation: x64 architecture and old x86. The first one benefits, but we still have choice between 64bit and 32bit software, we still have CPUs that support both set of instructions.
Why?
BACKWARD-compatibility while migration process. Yes, it takes long time, but there is no other choice. Ofc, you can teach child to swim by throwing him into deep water, but there is too much chances he can die and there is exist different, more successful methods.

Quote:
As a sidenote, mIRC is not forcing you to immediately upgrade to 7.1 as soon as it is available. 7.x has been put out there as a beta to show people how mIRC will be working.

That's why I reported about the issue (yes, issue, it's an issue for me and in this case I dont't care about your opinion, sorry; no offence, just the fact). I hope it will be fixed while new mIRC is still beta. That's the purpose of public beta-test. PPl try it, ppl see issue, ppl report.

Quote:
This gives your network admins time to start improving their Unicode support. They may not be done by 7.1, but you can continue to use 6.35 and tell them to start taking Unicode more seriously.

RusNet can't implement many important changes in their ircd since.. oh well, 2000? Not even talking about UTF-8 support. Same applies for DALNet.RU or many other big networks. It takes too much time and too complicated. Also, consider the fact IRC is almost dead and not so many ppl now care about it... oh well, even if it will be implemented - not before 15.x release.

Quote:
You shouldn't be complaining about the lack of support for codepages here-- you should be complaining about the lack of support for UTF-8 on your network.

ORLY.
I do.
I do complain like... few years already. But I'm very practical and rational - implementing backward-compatibility in mIRC "solves" this issue much faster.
Actually, the way how 6.x branch does it seems perfect to me (i mean the whole conception of supporting locale codepage and having UTF/unicode support at the same time; not talking about the issues of 6.x).

Quote:
If you took the time to write them the same kind of posts you wrote here

You have no idea...

--------

As I said already: I'm very practical and rational, the shortway is always better.

Also, I'd like to see Khaled's reply. Hope he will notice this therad >_>

Last edited by Lusht; 05/06/10 11:03 PM.

"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Oct 2003
Posts: 3,918
A
Hoopy frood
Offline
Hoopy frood
A
Joined: Oct 2003
Posts: 3,918
As I mentioned- Khaled has replied to this issue before (once in the very thread you linked to in your initial post). His stance is the same each time.

But to reply to the issues the ircd devs raised:

Nobody is saying that the server *must* do conversion for user data. If you encounter another client using CP you can simply tell them to switch to Unicode. mIRC 6.35 supports unicode, as well as every IRC client ever. The issue you reported had to do with network services and network messages using extremely old encodings that no sane OS or application uses anymore. They don't need to implement any form of encoding conversion or touch iconv at all-- they simply need to provide their server messages in a more appropriate encoding for 2010 (yes, it's 2010). If they can't do this-- well, they shouldn't really be hosting an IRC server.

mIRC 6.35 was the "backwards compatibility" phase of your x86 analogy. It supported codepages and Unicode alike. That phase is over and nobody moved on. Following on that analogy-- we're in a 64bit-only world now and those ircd developers still haven't rewritten their applications. Following your analogy further, 32bit compatibility layers in x86 arch won't stick around forever. Eventually 32bit code won't run anymore, just like eventually CP encodings won't work anymore. We're not there *YET* (mIRC 7 is still a beta), but it's coming.

You reported an issue, but you didn't report a bug. As the beta page states, the lack of CP support is by design. This release is to inform developers to update their code-- not the other way around.

PS. IRC isn't dead.


- argv[0] on EFnet #mIRC
- "Life is a pointer to an integer without a cast"
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Quote:
If they can't do this-- well, they shouldn't really be hosting an IRC server.

Don't tell me, tell them.

Quote:
But to reply to the issues the ircd devs raised:

I don't care. I came here to talk about mIRC and it's issues/features. Believe me, I had so much this bullsh*t discussions about IRCd and don't want to hear more.

Quote:
As I mentioned- Khaled has replied to this issue before

You said "it will be implemented", but I wasn't able to find where Khaled said exactly such. Also, I'd like to know when it will be implemented, at last approximately. Some of further betas or may be even 7.1 or 7.5 or whatever?

P.S. It is, I'm sorry :<

Last edited by Lusht; 05/06/10 11:20 PM.

"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Oct 2003
Posts: 3,918
A
Hoopy frood
Offline
Hoopy frood
A
Joined: Oct 2003
Posts: 3,918
Originally Posted By: Lusht
P.S. It is, I'm sorry :<


There's a good thread debunking this myth here.

Your networks might be dying-- and well, I guess that's a good thing, since your networks don't seem to support sane encodings anyway. IRC as a whole is not dying, though.


- argv[0] on EFnet #mIRC
- "Life is a pointer to an integer without a cast"
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Originally Posted By: argv0
Originally Posted By: Lusht
P.S. It is, I'm sorry :<


There's a good thread debunking this myth here.

Your networks might be dying-- and well, I guess that's a good thing, since your networks don't seem to support sane encodings anyway. IRC as a whole is not dying, though.


Sorry, the actual question was:
Quote:

You said "it will be implemented", but I wasn't able to find where Khaled said exactly such. Also, I'd like to know when it will be implemented, at last approximately. Some of further betas or may be even 7.1 or 7.5 or whatever?

I find this more important.


"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Dec 2002
Posts: 5,254
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 5,254
Unfortunately it is not possible to support both UTF-8 and codepages at the same time without compromising the way encodings are handled eg. mIRC would have to use a hybrid UTF-8 codepage combination that would not be valid UTF-8.

mIRC v6.35 did attempt to handle multiple encodings, using some rather messy kludges, and because of that users constantly reported issues with corrupted text.

The issue with codepages is most obvious in the channels list window. The channels list window lists thousands of channel names, many of which use codepage-specific encodings. However the window can only display text using one codepage. That means that channel names not using that specific codepage will be displayed incorrectly. The same issue applies to all channel and query windows where users chat in multiple languages.

While it may seem that mIRC v6.35 handles codepages correctly, there are a large number of situations where it cannot do so due to a lack of context, just like in the channels list window.

I realize that many channels use codepages and that this will be an issue for them. However, there needs to be a transition to UTF-8. The first step was to make mIRC v6.35 display UTF-8 by default. The next step, as far as I can see, is to make mIRC send UTF-8 text consistently.

Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Yeah, but channel list isn't so important. At last, if you're looking for specified channel with topic on specified language written with specific codepage symbols - you can set up this codepage in your system. mIRC 6.35 has great feature - it allows me to set up codepage per channel, like, for some international channels I use UTF-8, for most others I just do not use any UTF-encoding and mIRC handles them properly, relying on my system codepage. Which is great and I don't care about other languages (especially in channel list windows) since mostly I need cp1251 + UTF. Every international channel uses UTF-8, because there is no IRC-client that supports all codepages at the same time for same windows and - tbh - such ability looks like utopia, lol (and make no sense too).

Why just don't add support at last for system's default codepage somehow? For example, like in other clients where you can choose codepage when select a server profile.

Last edited by Lusht; 17/06/10 07:36 PM.

"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Oct 2003
Posts: 3,918
A
Hoopy frood
Offline
Hoopy frood
A
Joined: Oct 2003
Posts: 3,918
Originally Posted By: Lusht
Every international channel uses UTF-8,


Then why do you need codepage support?


- argv[0] on EFnet #mIRC
- "Life is a pointer to an integer without a cast"
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
...
I said international, not national...


"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Oct 2004
Posts: 8,330
Hoopy frood
Offline
Hoopy frood
Joined: Oct 2004
Posts: 8,330
I think it's time the national servers follow the international servers then. There's no good reason to use code pages when unicode has been around for so long and is so widely used these days. From all the posts I've seen from Khaled on this issue, I doubt you'll see code page support in mIRC 7.x. If your server(s) can't use unicode/utf-8, then it looks like you'll have to stick to 6.35. Keep in mind that most of what mIRC 7.x currently has to offer isn't going to help you anyhow if you're not using unicode.


Invision Support
#Invision on irc.irchighway.net
Joined: Dec 2002
Posts: 5,254
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 5,254
I understand what you are saying, however I think it is important that IRC makes the transition to UTF-8. If the new version of mIRC adds support for codepages then IRC users will continue to be stuck with codepages, along with their limitations and broken support when mixed with other encodings, for a very long time. IRC needs to move forward. Hopefully by making the new version of mIRC use UTF-8 consistently, and with other clients doing the same, IRC as a whole will slowly converge towards UTF-8 as standard.

Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Ok, let's forget about usability code[age support in 7.x, but what's about some backports for 6.35 users? At last performance & compativility fixes for Windows 7.
Srsly, the only version that works decent on Win7 is 7.x but it is impossible to chat with it for non-english/non-latin users.
That's a bit sad.
You must understand, it is not just me, many users use such kind of codepages and they're doomed to stuck with 6.35 (there is already some threads about this situation in this section). I'm sure, they would appreciate some backport fixes for 6.35.

Is it that hard to do?

Last edited by Lusht; 19/06/10 10:38 AM.

"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Joined: Dec 2002
Posts: 5,254
Hoopy frood
Offline
Hoopy frood
Joined: Dec 2002
Posts: 5,254
Users have often requested this for previous versions of mIRC, however it is not easy to backport when there are hundreds of subtle internal changes between version releases. In the case of v7.x it would be even more difficult since porting mIRC to Unicode has required over a year of work and tens of thousands of changes to the source code. It would be incredibly time-consuming to track down the differences in the source code between v6.35 and v7.x to see where changes and fixes were made that could be applied to v6.35. The versions.txt file for v7.x only lists two to three hundred feature-related changes, however the actual number of internal changes is enormous.

Apart from that, I put a huge amount of work into each new version of mIRC and I just do not have the energy or time to then go back to older versions and put in a similar amount of work into them all over again.

It would be much easier to implement codepage support in v7.x (or at least I think it would be, I have not looked into it. Since mIRC is fully Unicode now, there may be side-effects or issues that are not immediately apparent), which is something I would rather not do at this time for the reasons I have listed in my previous posts.

Joined: Oct 2004
Posts: 8,330
Hoopy frood
Offline
Hoopy frood
Joined: Oct 2004
Posts: 8,330
mIRC 6.35 works just fine in Windows 7. The only thing that doesn't work is $os because it doesn't know what OS you have and returns Vista. Other than that, it works as well in Windows 7 and it does in Vista or XP. I've used 6.35 with Windows 7 every since Windows 7 came out and I still use it for my trivia bot. If you have specific issues with 6.35 in Windows 7, you should bring them up in case they need to be fixed, but otherwise, it works just fine.


Invision Support
#Invision on irc.irchighway.net
Joined: Aug 2007
Posts: 42
L
Lusht Offline OP
Ameglian cow
OP Offline
Ameglian cow
L
Joined: Aug 2007
Posts: 42
Khaled
Quote:
It would be much easier to implement codepage support in v7.x (or at least I think it would be, I have not looked into it. Since mIRC is fully Unicode now, there may be side-effects or issues that are not immediately apparent), which is something I would rather not do at this time for the reasons I have listed in my previous posts.

Then, may be, at last some separate builds with codepage support when stable version of 7.x is out? Hell, let it be even highly experimental, I don't care while it understand how to work on my "native codepage" and does not lags on win7.

Riamus2
Yeah, right
go tell me how wonderful it works on win7
I'm so clueless and full of lie here...


"Do, or do not. There is no 'try'." - Yoda ('The Empire Strikes Back')
Page 1 of 2 1 2

Link Copied to Clipboard