What Now ?
Looking back I find I started this blog on September 24th 2010 just short of a year ago. The goal was to document the journey to Tarrasch V2. That journey is now at an end. What is more important, the journey or the destination ? For long periods of the journey I was exasperated with it, and worried it would never end. How could I ever finish this thing ? But now I have “arrived”. For a very short time I felt elated, but now I will admit to a feeling of flatness. What now ?
Initial reaction has certainly been muted. Not a lot of buzz in the comment section of this blog. Download stats show an initial burst leveling off fairly quickly. Just the same as for both beta versions. Just at the moment the danger is that Tarrasch V2 is just going to replace Tarrasch V1, without becoming significantly more popular (basically the background level of activity is around 30 V1 downloads a day).
I strongly suspect the problem here is promotion. Tarrasch V2 retains all the ease-of-use benefits of Tarrasch V1, and at the same time is vastly more functional. Objectively speaking, it seems a no-brainer that it should take Tarrasch to a new level in popularity. But for that to happen people need to know it exists, and there isn’t a mechanism for that in place. Downloads tend to come from a set of existing links, none of which are “V2 aware”, or from people googling for Tarrasch Chess GUI, i.e. they’ve already heard about Tarrasch. That is not likely to be a “V2 aware” source either.
I suppose I can choose to either “let nature take its course” and see whether Tarrasch V2 slowly picks up momentum (or not), or I can do some active promotion.
I started this post intending to write about the future of Tarrasch. Should I repeat the V1->V2 journey with a similar V2->V3 journey ? Or is it time to adopt a more mature process, making smaller, more incremental improvements, so V2.1, V2.2 etc. Or maybe I should simply declare victory and stop.
I will ruminate on these unanswered questions.
The Tarrasch Chess GUI 
‘Or is it time to adopt a more mature process, making smaller, more incremental improvements, so V2.1, V2.2 etc.’
Yep, I think that’s the way to go!
Bill, Tarrasch is excellent but it’s marketing is insufficient or non-existing. I only know Tarrasch because I’m a power user. Regular and casual users wouldn’t have a chance, but they are your primary audience. This is really difficult, because you don’t make any financial benefits. You wouldn’t spend countless hours in Facebook promoting, you wouldn’t do that in Twitter as well. The bottom line is, the new release is good but it will never be popular.
It will never be popular ? Ouch.
This is life. Without corporate support, it’s very difficult or maybe impossible. Chessbase can do it because they are huge. Try to post more to discussion boards?
Don’t listen to him. There is many Open-Source projects which do not have advertisement but they are still popular. Currently if you Googe Tarrasch, the main page of the website will be the first one to appear. I think there is many more downloads then comments. Do you have like a download meter to see how many comments are there?
PS: I am really waiting for v2 source code to be released!
Saying it will never be popular is pretty negative thinking. There is no way you can know something like that unless you have a mystical crystal ball. It might very well catch on.
I don’t know where you have promoted it so far, but as adsoad mentioned, discusiion boards would be a good way to go if you have not yet. Start threads on sites like Chess.com to answer peoples questions and to promote at the same time. Maybe start a thread asking people to submit there own commented list of move games. You could put them in a download section here on the site.
Update the Tarrasch GUI front page. It’s pretty vague about what all Tarrasch can do exactly and who it’s targeted towards. Put up more screenshots that show different features in 2.0. Maybe make a small guide for the program.
Put Tarrasch on various download sits like download.com, fileforum, and softpedia if it’s not already.
Send a note describing Tarrasch to Alexandra Kosteniuk.
She might post it on her blog – which is very popular.
Hello! I downloaded Tarrasch source code and tried to compile Toy Engine (since I don’t have WxWidgets installed on my Windows machine) with VS. I got several errors:
Error 3 error C2365: ‘log’ : redefinition; previous definition was ‘function’ ..\tarrasch-source\ToyEngine.cpp 22 1 ToyEngine
Error 5 error C2451: conditional expression of type ‘overloaded-function’ is illegal ..\tarrasch-source\ToyEngine.cpp 40 1 ToyEngine
and several more… but they all seemed to come from FILE *log.
I don’t usually use Visual Studio. I prefer Code:Blocks, with GCC. It is much easier for me to find errors there, and it just feels like “home” :D. So I decided to try to port Toy Engine to Code::Blocks. So I made a Code::Blocks project, copied all the files, and got ready for long hours of changing Visual Studio specific code to standard C++. Surprisingly, all I found was that inside your classes you did something like this:
class d {
int d::function() {return 0;}
}
instead of
class d {
int function() {return 0;}
}
and GCC doesn’t allow that if the function declaration is inside a class! So I changed it a bit, which took me about 5 minutes, and it compiled!
Now, I will try to port the Toy Engine on Linux (first on Ubuntu with Code::Blocks, and maybe later with makefiles), and will let you know of the results.
Then I want to get the Wx installed on Windows, and port the GUI Tarrasch itself from Visual Studio to Code::Blocks, and then maybe to Linux 🙂
Then I will try to understand the code of the project itself a bit more, and maybe contribute some code. I also think, you should be able to get more developers (and users) for this project by making a sourceforge page, and maybe changing the licence (because your current one seems to be a bit loose, and easy to misinterpret (?)) to GPL, LGPL, or Apache. More developers will cause there to be more talk about this great project, therefore the number of users should increase greatly.
I would also really waiting for the second version of Tarrasch source code to be released!
Kostya, thanks for your interest and your kind words. I am sure you are right that a Sourceforge/Github page and a standard licence are the way to go if I want Tarrasch to be a real open source project. My apologies that your experience with the V1 source code was (much) less than perfect. The code quality is a bit uneven and the toy engine code in particular is unloved. Basically, I was just hacking at that until it worked ! Yes, I recall I had some instances of classname::methodname inside the class definition which is unnecessary/wrong, but as you say Visual C++ is quite happy so I didn’t necessarily fix all such instances. Sorry! The toy engine is a sort of project within a project. It was written just to gain experience with the UCI protocol and also to allow me to include a simple engine with the GUI. At that stage I didn’t realise how easy it would be to get authors of “real” engines to give me permission to include those engines!
I have been very busy (with “real” work) for the weeks since finally releasing Tarrasch V2, and that explains why I haven’t been very visible. Soon I will be back and I will address the issue of the missing V2 source code. Time spent on understanding the V1 source code will not be wasted if you confine yourself mainly to the pure chess classes, they haven’t changed much at all but the GUI presentation code has obviously changed quite a lot.
My experience with Tarrasch is not “much” less than perfect. It is real life program, and if working with real life code, probability is that there is going to be errors and hardships. This is not worse then many other majority of other projects.
Anyway, I was trying to port Toy Engine on Linux, and everything was going fine, until I encountered Semaphore and windows threading :). It may be a bit of a challenge to port it due to the hackishness of the source-code and my inexperienceness. But that’s the whole point of programming 🙂 I also spend several hours trying to understand the source code, and it all finally starts to come together in my head. I also ordered a book on Wx, so I can contribute and understand the code better. This is really exciting 🙂
I found Tarrasch GUI through downloading the Rybka 2.3.2a demo which had v1 and now has version 2 Tarrasch front-end. I was impressed and donated some money to help or at least show appreciation to Bill for his v1 to v2 journey.
During Bill’s journey, I no longer use a PC, I switched to a mac to use some mac only software. I use Tarrasch via wineskin as its own app, so it is cool to still be able to run this software on a mac, especially good since I can run various open source chess engines such as stockfish, houdini 1.5 (pre commercial release), komodo 3 and critter 1.2.
Stockfish had a GUI for mac, none for PC so in my opinion, I think a hook up to the chess engine authors maybe a good way to get attention.
Yes, we can use arena but Tarrasch is compact.
I haven’t downloaded the program yet, but will when I get home today. Just stumbled accross it when I searched for advice on how to play the Tarrasch opening. I agree that marketing is probably the issue though. I have never heard of this until now. True, when you search for Tarrasch, you find a link to this page, but if you search for “free chess program” it’s nowhere to be found. I would hit up a couple daily bloggers on Chess.com and see if anyone will mention your program there. Maybe post a video on youtube or ask kingcrusher or someone like that, who posts chess videos frequently, to mention or demonstrate your GUI. Good luck, and thanks for puting in all the hours.
Thanks for this thoughtful comment. No google hits on “free chess program” is obviously a problem. I had no idea! Please let me know, here or by email, what your impressions of the program are once you’ve downloaded it.
I am tied up left and right with other things at the moment, but soon I will be spending some more time on Tarrasch, and for the first time really I will be putting some effort into promotion. I will definitely come back and reread your comment (and the other comments here – thanks all) for ideas.
Well, I haven’t been able to look at it as much as I’d like, but I love the simplicity of it. It is extremely user friendly. Great work, I say. Good luck on getting this thing promoted. I will definetly recommend it to others. Thanks!
Well, I first found out about tarrasch since the komodo engine has a download with their engine plus tarrasch, and I love it. I like the simplicity behind it, especially since alot of free engines have autorotation of the board when playing black. The only suggestion I have is to have a clock option for number of moves for a given time limit. If this is already implemented then I apologize for that. Either way, keep at it, I for one am a loyal tarrasch user.
Until this evening I had not even heard of TarraschGUI. I was reading the documentation accompanying the Glass engine and your GUI was mentioned in the introduction of the document. When I made that discovery I promptly downloaded it.
It’s a good package, worth a download. It’s unlikely to get noticed, though, unless you find some way to push it. “Free” is not synonymous with “well-known”.
For future versions I would like to make a couple of suggestions. 1) Consider making a table for engines so that it does not require hunting the file system every time you want to change engines. 2) It could use proper documentation, like a real help system instead of a dialog box, or even just a PDF in the application directory.
Upload the program for free download on Chess.com! People will find and use it!
I’ve tried quite a few chess GUIs and ended up with Tarrasch v1 about a year ago, and now moved to v2 (which evolved in a right way). The best thing you can do to make it popular is to contact opensource chess engine developers — for example I downloaded Stockfish engine a few days ago, and on their page they only give links to ChessBase (didn’t try it) and Arena (imho it’s ugly) UIs.
I also would like to suggest a couple of features: 1) Flip board when setting up a position. I play a lot on gameknot.com and when you play black there you can’t flip the board, so setting up a position for analysis is annoying. 2) Adjustable board/piece size. Or if it’s already there, it should be more obvious 🙂 I also wanted to complain about absence of analysing mode, but found it just today — for a non-native English speaker something like “Analyze” / “Analyze mode” will be much more preferable. “Kibitzing” really sounds like some of those crazy variations of chess.
And, of course, thank you very much for all your efforts and time spent to develop this amazing program!
Hiyah, the thing is, coming here from Komodo’s website, I do not quite understand what this GUI does that for example Shredder does not do. I was looking on the main download website (triple happy) for a list of features and possible uses, but that was really difficult to find. Maybe it is somewhere, but in that case it is not so obvious where. Why not say what it is for: an engine interface? A database? A training tool? If so, what kind of training (memorization of openings?, find best next move?). Most people, including me, are quite happy with their chess training set ups, so saying “just try it” will not cut it. There must be a good reason for it.
Thank you. This is harsh criticism, but fair I think and something of a wake-up call. Ironically I personally am regularly infuriated by websites that assume prior knowledge and don’t explain their purpose in a clear, logical manner. It is disturbing to me that I have apparently made this mistake myself.
I’ll quickly address your key concern in this post. Tarrasch is a chess GUI (graphical user interface). Without a GUI a chess engine has no way of interacting with users. There are other chess GUIs out there, Tarrasch attempts to differentiate itself by extreme ease of use. Tarrasch seeks to make it as easy as possible to perform a few simple but important chess activities on your Windows PC. 1) Play a game against a chess engine, with either colour, not necessarily from the starting position, with a time handicap if desired. 2) Setup an arbitrary position and get the engine to analyse that position, view and play through the analysis provided by the engine. 3) Create and edit .pgn files. Enter variations easily (just make moves anywhere). Enter and edit comments easily (just click anywhere and start typing), Promote and demote variations and even comments (so comments can change to moves and vice-versa – great for picking text off a website say and converting it to .pgn).
I don’t think any GUI makes these tasks as simple and intuitive as Tarrasch. I could be wrong. I am biased. If as you say you are happy with your existing training setup, maybe Tarrasch is not for you. Tarrasch is for people who think other chess software is unnecessarily complicated and confusing. Tarrasch is great for beginners, but some expert users like its approach and use it a lot as well. Personally I use it for everything except serious opening preparation, because it does not have the database features needed for that.
Tarrasch has a few novel and interesting features, these can all be discovered by exploring the menus after mastering the basics that I have already described. But as you have pointed out, this is not a satisfactory situation. I need to update the website which has lost focus. Tarrasch has been revamped in the last year, but the website has not. In particular I removed the FAQ (which may have answered some of your questions), because it matched the old version of Tarrasch. I promised to replace it with a new FAQ, but have been tardy. I will get on with it!
One more thing, if you are still interested in Tarrasch, the first two posts in this blog (accessible from “About”) are a good place to start.
Great, thanks, I think this comment would make a great start for an introduction of what this GUI is about. It seems that a mix of this + Chess Hero would be really quite nice, as Chess Hero does not make annotations easy, but is great for trying out variations as well.