You can use the server-side deck ID to prove you're always playing the same deck.
A feature to share the server-stored decks with other people will be made available eventually. But in the meantime, if the tournament organizer sets up his own Cockatrice server, checking the deck lists is easy by extracting them from the database by ID.

A new release is out. Link to website
 Changes:
 - Russian translation by Alexander Davidov
 - Countermeasures against chat flood, ability to kick people
 - Buddy list, ignore list
 - HQ picture autodownload by Sergey Klimov
 - improved sideboarding view
 - added 'undo last draw' feature
 and other things.
 Be sure to run the Oracle tool after the update as the file format has changed.
 Have fun!

You need to download the current version, and apart from that, it also means that there'll be a more readable error message in the next release.
MBS has been in Cockatrice since January 28th
You just need to re-run the Oracle tool.

Yes. Sadly, I haven't had time to incorporate your ideas about the build system, but it's going to happen eventually.

Again, it's time for a new release. As usual, the client can be downloaded on the Download page.
 Since it's been a long time since the last release, there have been a couple of changes:
 - French and Japanese translation
 - some visual improvements
 - smoother picture caching
 - cards in zone views can be selected
 - removed chat rooms since they were not used, implemented play rooms instead
 - changing the controller of cards works
 - removed economical table layout, implemented stacking instead
 - and a long list of minor fixes, see for yourself.
 
 Attention: The main game server has moved to another IP address. A DNS update is taking place, but it might take a day or two to finish. In the meantime, if you cannot reach the new server via its DNS name, please use this address: 78.47.181.111

I don't know how or why it works, but there are in fact draft sites that work, so the only real challenge is to find enough people willing to play a couple of rounds with the drafted decks in a reasonable manner. Apart from that, it shouldn't be a problem drafting on one of those sites and importing the decks into Cockatrice. I do it pretty often, though I don't know enough drafters that well, so we always have to use a couple bots, which makes the draft decks unrealistic. That is not a fundamental problem though.

Thank you, and also for the article you published. It's very detailed and well written.

I have thought a lot about this question as well and I wanted to do it. It would facilitate finding people to draft AND play very much, as otherwise there is no connection between the draft service (netdraft etc) and the game server.
However, I don't know if it is legal to do so. The game server would have to contain a card database (which at the moment it does not) and transmit some of the contents (card names, at least) of this database to the clients. IANAL, but I believe this information is copyrighted and it could be a large risk to set up a game server that would offer this service.
Of course, it wouldn't hurt to implement support for it protocol-wise, but as long as it's not clear whether the support can be used at all, I don't see a reason to.

Quoted from the Cockatrice website:

The server crash issue has been fixed. This is not to say there aren't any other bugs in the server, but we'll see what happens.
I would think that this happening without your opponent being dead is kind of unusual In situations like this, I often make like 3-5 tokens (so that the opponent can block and play removal) and put a counter with the appropriate number on another one. That way, the play area doesn't get cluttered up. Stacking those tokens (or any creatures with the name name, for that matter) to save space might be possible, but I'll have to think about it for a while.
Agreed, though you can press Ctrl+G to make the same token again.

I apologize for the server crash that happened today. This issue will be resolved, but it's going to take some time as I'm busy at the moment.
You can attach a card to another. This is meant for auras, equipments and the like. Why would you like to do that for cards that don't have a relationship?
The context menu is supposed to work just like it does on the table, though I realize that you can't select multiple cards at once in the library viewer. Scrying and similar effects are a real pain. This will hopefully be implemented soon.
This discussion has been brought up a lot of times, and I'm going to think of a better solution eventually. However, this requires a lot of testing and shouldn't be changed hastily.
Also, thanks for correcting my linguistic mistake (economical vs. economic). If anybody notices a mistake like that, please don't secretly laugh at me, but report it instead so that it can be fixed. This also applies to everything that, while technically correct, nobody would say like that.
It's implemented, it just requires a correct SQL table setup. I've added a file called servatrice.sql to the git repository just now. You can use this file to create the correct structures, though be aware that I'm going to change some of them eventually.
In servatrice.ini, you need to set [authentication] method=sql and [database] type=mysql (with the right login data, of course). Restart the server app if you manually make changes to the table data.
It probably will. The current version has at least one bug causing a crash. If you could make a core dump when it happens, that'd be helpful

Measured by the number of users on the server in the past few weeks, it seems the community has been growing continuously, which is very good. Hopefully we'll be able to resolve the 'critical mass' issue soon enough.
Inspired by the fact that Cockatrice user Gocho has just contributed a Spanish translation file, I'd like to ask if there are any volunteers out there to translate the user interface to other languages (at the moment, only English, German and Spanish are implemented). You don't need to know anything about programming, there is a simple to use tool for it, and I will explain the process to anybody who is interested. The only problem would be that sometimes, minor updates are required to keep up with the development of the UI.

This has obviously been bugging me as well. All the 'reveal' features (library, hand, single cards) will be implemented at the same time. I just need to think of a good user interface for it.

Quoted from the Cockatrice website.

That would work. Not having to hold Tab and seeing the information anyway works even better.
This comment only shows that you've never really used the program. When you click a card that is in your hand, it will fall onto the table at a predetermined position. I never drag a card from my hand anywhere. The place where the card is put is determined by its card type. The "stack" row isn't really a stack but a row reserved for instants and sorceries. The differences are that both players have such a row, and that there are no virtual objects (abilities, copies etc) to be put there. As I said, it's not possible to create a separate zone for non-permanent spells without forcing users to always single-click the cards into that zone (instead of dragging). If I were to do that, I would have to implement some other things as well, mostly dynamic creation of temporary piles for cards like Gifts Ungiven or whatever, since you wouldn't be able to just drag the cards onto the table any more. The idea is not really bad, I just don't think it'll be easier to use than it is now, so all that has to be considered very carefully.
In the beginning, there were no rows, and you could put cards wherever you wanted. This was a bad concept because it implied a direct connection between rendering and server-side coordinates. Using the concept of a fixed table grid that is not directly correlated with display coordinates, it's possible to display the pieces of information given to you by the server the way you want it. The option "economic land layout" is an example of this. Keep in mind that the rows need not necessarily be displayed as rows, but the cards in one row share the same logical row index to group cards that contextually belong together (lands in one logical row, creatures in another, instants and sorceries in yet another).
I'm not going to discuss this any further as long as you're not willing to understand what this is all about. Of course it's not optimal, and there is a lot of room for improvement, but you're thinking in the wrong direction here.

The message log is not only a chat box, but it shows all changes to game, and it does it better than the table itself does. While playing, I pretty much look only at the chat window while my opponent does stuff. This way, you can process the change in information much faster. This is because there are no hands moving cards around like there are in real life, so changes happen instantly, and you don't know where to look for them at a given moment.
It's part of the game state. I think every non-private part of the game state should be visible at all times. From the game's perspective, this includes the number of spectators and the deck IDs of the players.
This is already the case.
This is because in real life, people don't save a row for spells on the stack. I had been considering making a real stack for a long time, but ultimately decided it complicates playing a lot for very little benefit. Again, the biggest problem is that many people don't use the predefined rows like they're supposed to (it would even save them many clicks) but throw their cards on the table anywhere they like. Because of that, the "stack" has to stay on the table and cannot be put into another zone that would dynamically disappear when not needed. This is what I meant when I said it can't be done without implementing the game rules.

I might agree with that.
In fact, I do, just obviously not a professional one.
I'll forward these ideas. Sadly I don't have time to learn how to do that ATM.
Which one do you mean? Apart from the solid color, Cockatrice comes with three different textures for the play zones. I like the "fabric" one best.

In a tournament setting, if for some reason I need to know the number of cards, I count them. Since you can't do that yourself in such a program, the number has to be displayed somehow. Why make the user waste mouse clicks when the number can be displayed right there on the pile where no other useful information could be displayed anyway?
Voice chat itself is not nonsense, implementing it in such an application is. There are perfectly working implementations out there, and I don't see a reason to reinvent the wheel. Of course, it would be nice to have some kind of integration, but I'm not going to write everything from scratch just for that purpose.
It can, it's just not necessary.
There's going to be some kind of deck list sharing mechanism. Since for tournaments, the server-side deck storage tool will have to be used and the decks will be stored in the database, it won't be difficult to publish them in some way. I'll have to think about all that very carefully, though.
Any analysis you want to do, you can perfectly do with a suited analysis tool. This program is for playing, and overloading it with features only very few people need will not exactly help beginners understand it.
You may be right about that, but adjusting this to look perfect could be quite a bit of work that I don't have the resources for.
Are you saying that everything that behaves like a button should be rendered with that beveled gray corner? I don't think the mana counters would look very good if they were rendered like that. If you don't mean that, I apparently don't know what a button is.
You can do with far fewer mouse clicks if you just let the server choose the card. Your randomly discarding a card has nothing to do with your opponent, and it shouldn't have.

It's not my fault if you didn't read what I wrote on the subject before. The table has four rows, and users who like to place the cards freely want to have those rows. If everybody just let the cards fall into place by single- (or double-)clicking, empty rows could be hidden. As this is not the case, they can't. So the height of the table is 8 cards total, and there's nothing you can do about that. Your idea to make the hands horizontal (which is possible through configuration, by the way!) takes away even more vertical space.

My laptop has a touchpad. I don't know if I'm living in some kind of Neanderthal world there.

Fine. I'm not saying I won't do it


Are we talking about the same version of the program? I don't have a black area left of the play field. Like I said, the green one can be transferred to the bottom middle by clicking a checkbox in the Apperance tab of the config thing (if you'd like to have your cards displayed even smaller, that is).
I agree, it's ugly. The background is not green, but rather a lighter shade of grey than the rest. I don't know how to make it look nicer.
Yes.
Thanks. Honest and constructive comments I always appreciate, critism due to lack of understanding of the problem at hand too high a level of abstraction, I do not so much. Sometimes it's difficult arguing with you as your posts tend to contain a little of both.
MWS sure is a masterful piece of eye candy

I agree.
I'm not developing for Mac specifically, I have OSX for the sole purpose of compiling this application. If it needs to look perfect on one platform and there's only one way of it actually doing so, it should look the same on all platforms, right? And with some exceptions (fonts etc.) it pretty much does.
Let's not let this thread derail into some kind of OS / GUI toolkit bashing discussion.

While this may be true, it would be more constructive to come up with an idea how it should be done instead.

Why don't you load one? I don't see a reason to include something special like that in such a general purpose application.

No, all this is nonsense.

I don't use any fixed width fonts in the interface. This might be a Mac issue I haven't noticed.

Why should they be buttons? You don't have buttons in real life either.

This I get (and it seems reasonable), but
this I don't.

Not doable any other way without implementing the game rules. Already explained A LOT of times.

Where would you like his hand to be? I could see myself hiding it altogether since it only displays blanks anyway.

The battlefield cannot be made readable, see above. An MTGO-like middle-click zoom could (and probably will) be implemented, but what will touchpad users do who don't have a middle button?

Yes.

Possibly. Would be the first thing I'd turn off though.

As is the case throughout your entire post, I don't have a clue about what you exactly want. This doesn't help at all.

Yes, I just haven't had an opportunity to actually restart it as there were always people connected to it. Now it should be fixed.

The instability has been fixed. I'm going to update the server as soon as there's no game running.

You shouldn't be, discovering bugs is a good thing. I think we're going to see at least one extra crash (hopefully with log this time) before it can be fixed though.

Apparently, you crashed the server. This needs to be fixed ASAP. Do you remember what your last action was (and in what kind of situation)? I'm afraid I haven't got a log file because it accidentally got overwritten

All this means is that the MWS developers didn't have a clue about what security means. If, hypothetically, someone were to successfully reverse engineer the protocol, all these assumptions would go down the toilet.
Yes.
Does that mean it's actually still being developed? BTW, I think a transition from Delphi to Qt requires a manual rewrite of A LOT of code.

New release! Quoted from the Cockatrice website:
This version breaks protocol compatibility, so you need it in order to log in to the current server. Be sure to re-run the oracle tool as well as there have been some changes there, most notably the renaming of the cards starting with an AE ligature to just Ae so you can write the name on a typical keyboard.

This sounds like a sensible thing to do. I tried to reverse engineer the MWS protocol when I first started the project, but it didn't matter enough to me so I abandoned it after seeing it wasn't even remotely a human-readable protocol. Of course, it would be really good if an interface between the protocols could be made.

Thanks. Maybe once we reach beta stage and have proper versioning, we can get into other distros as well...

This idea is certainly good, however I don't know if it can be done (and if it can, I don't know how). As far as I know, the MWS protocol is not known, so it would have to be reverse engineered.

Strangely, I had written a reply to Rich's post, but it's not there. Maybe I forgot to actually click OK.
Anyway, what you want to do is not possible with the current version (only if you open two instances of the client and a server on your machine, which is possible, but in no way fun). Since there's only very little work left to do to make that work, I hope it'll be in the next version. As of today, auras/equipments work, as does targeting a player with arrows.

Because of a strange bug (or feature?) in Gatherer, I thought they just hadn't put up the M11 cards, while the cards were in fact there and were just not displayed if your browser language wasn't set to English. Now that I know, I've added M11 to the sets file that lies on the web server. The new cards can be retrieved via a full oracle import.

Duplicating a card is already working in my git branch
The parsing thing requires a bit of deliberation as it pushes us very much into a Magic specific direction while also being sensitive to even small oracle text changes. I think that if we were going to do something like that, it would be better to implement the entire Magic rules system so we'd be getting features like the one you're asking for as a side product. This isn't your first idea pointing towards that, so would you be interested in doing it? I find the idea very promising, but I'm also aware of the amount of work that would have to be invested, and I don't have the time to do it.

This would give you additional information without your opponent knowing about it. You see, if you drag the top card of your deck onto the table face down for some reason, you shouldn't be allowed to know what it is (except if you explicitly look what it is, in which case your opponent should be informed about it). For that reason, the server doesn't even tell the client what the card is, so a query has to be issued anyway.

Thanks. The first idea is very good and has not been had before. The other things are actually already being worked on, especially the equipment thing. The problem how an equipped card should be displayed is not trivial, but I believe the solution is near

Too late, already coded
Do you think this is a matter of getting used to it, or could it actually be improved? Problem is, I don't have any idea how.
Yes. Also coded The code is not going to get into the git branch for some days though.

This has been reported many times, and I've never been happy with the idea of not being able to create something like a dredgeable Life from the Loam token (for example, when you cast Cunning/Burning Wish without having a sideboard). In my opinion, this inability has always been a major flaw in MWS. However, as everybody seems to want it, I've made a checkbox (which is checked by default) for it so you can now make tokens that disappear once they leave the table. Keep in mind that this doesn't work with the current version of the server as I don't want to force people to update for such a small change.