Author: enrico carrisco
Date: 01:46:18 09/21/05
Go up one level in this thread
On September 21, 2005 at 00:45:49, Axel Schumacher wrote: >On September 20, 2005 at 23:05:15, enrico carrisco wrote: > >>On September 20, 2005 at 19:29:20, Axel Schumacher wrote: >> >>>On September 20, 2005 at 18:29:06, enrico carrisco wrote: >>> >>>>Hello Axel. >>>> >>>>It appears you have spent a great amount of time on the review and such a >>>>commitment on time is a noble thing. >>> >>>Thanks. >>> >>>> >>>>I have a couple of questions about a few of the features the programs were >>>>ranked on as well as the overall evaluation of Palm HIARCS. >>>> >>>>Clearly aspects like board, piece design and a few other odd categories are of a >>>>subjective nature and what one person finds appealing the next could find >>>>unappealing. >>>> >>>>However, a few other categories caught my eye either because they did not make >>>>clear sense to me or were just plain erroneous as they apply to Palm HIARCS. >>>> >>>>1) Hashing - I'm having a hard time making the connection as to how this is >>>>relevant - surely what matters is chess strength, not that someone can allocate >>>>some memory! Maybe there should be a category for pawn hash table or memory for >>>>the move list!? >>> >>>Points for hashing come from different sources. First, if the program has >>>adjustable hasing in general, which is important IMHO, since the playing >>>strength depends on it highly. Yet, the on a device with high memory capacity, >>>let's say 264MB RAM, some programs will still use only 64k Hash. Second >>>adjustable Pawnhash and thrird other ways to adjust hashing (some programs offre >>>even more hash-functions) >>> >> >>Hashing is only a component just like many others which help chess strength, At >>the end of the day it is the chess strength, not the hashing that counts. >> >>If a chess program did not use hash tables but achieved the same strength as a >>program which used 16Mb of hash, by your measure, it would score less points. > >No. > >>However, a program which was as strong and used less memory would actually have >>some advantages. > >You are totally wrong here, sorry. The playing strength is strongly connected to >the hash settings. So, if you run a program (Hiarcs or any other) in the only >possible hash-setting (hidden to the user) it will have only a strength >independent of the hardware. It would make a HUGE difference, running a program >with 64k or with 128MB. The same thing is important the other way around. Maybe >a user want to decrease hash settings, because the program uses too much of it, >leaving less for other processees of the CPU. > > >> >>>> >>>>2) In the "Save Game" ranking, Palm HIARCS scored 4/10 - Last I checked, Palm >>>>HIARCS saved games fine. Is this not a "has it" or "doesn't have it" type of >>>>feature? >>>> >>> >>>No. >>>Hiarcs can only save very inconvenient via MemoPad and it is difficult to >>>select, for example a specific PGN database where to save to. There are many >>>different qualities in the ability to save a game. Also, for example, the >>>ability to comment during saving, choosing in which format to save etc.. >>> >> >>You already have separate categories for PGN support, selecting a database, > >that's a different thing and has nothing to do with saving > > text >>in saved pgn, > >again, that has nothing to do with saving > > >etc. -- so are you not double rewarding or penalizing programs? > >No, not at all. I'm simply very correct in the measurements. > >> >>By the way did you not consider the HED tool for importing PGN databases into >>HIARCS or the HRM tool for managing PGN databases? > >Are they Hiarcs products? > >> >>> >>>>3) In the "Level Adjustment" ranking, Palm HIARCS scored 6/10 but you can adjust >>>>more on Palm HIARCS than on Fritz 8! ?? >>> >>>No. Hiarcs offers 'only' 10 different playing levels. Of course, there are other >>>ways in Hiarcs to adjust playing strenght by other parameters, but then they are >>>covered in these other attributes categories (e.g. personalities, style, hashing >>>etc.... >> >>You are very wrong - it seems likely from this comment that you have never used >>a fully functional registered Palm HIARCS! Did you actually review the full >>product or just the free demo you (or your friend, as you say) downloaded? > >My friend had the full version. >regarding the 10 levels, I just had a look at the Demo, of course (I tested it >previously correctly). I don't know why you complain about 6 out of 10 points >for that category anyway. That is a very very good score. None of your close >competitors, neither Fritz, nor Chess Tiger, nor Pocket Grandmaster, nor Chess >Genius, nor ChessPartner, none! of them, scores as high as Hiarcs. > >> >>Sure the unregistered/limited features PH has only 10 levels but the activated >>product has much much more... >> >>It has 10 instant sparring levels beginner to challenger, it has loads ELO >>levels from 750 Elo to 2000 Elo - all highly realistic and full strength, not to >>mention all the time/level options for casual, game in, move bonus, time >>controls etc, etc. > >which are all covered in seperate sections > >> >>> >>>> >>>>4) Engine strength counts for ONLY 10/383 points - 2.6%?? I think engine >>>>strength is definitely relevant on such slow devices compared to PCs. >>> >>>Engine strength is valued a lot, next to design the highest! There are 100 >>>parameters.For 99% of the users, the strength is not of first priority, since >>>they will loose against the program anyway. >>> >> >>I think you are very wrong, because these devices do not pack the punch of a PC, >>chess strength of the software is VERY important. It is very useful for >>analyzing games, book moves, problems etc. Top strength also means you don't >>need to let it think so long to get a tough game which saves the battery life >>significantly! >> >>There is a big variation of chess strength on the Palm and Pocket PC - I prefer >>a chess program that can turn on full strength when needed, there is a big >>difference between a 2550+ Elo program and a 2200 Elo program when you are >>playing over a GM game for example. >> >>Not only that, Palm HIARCS is the only Palm chess program with wins over GMs - 3 >>different GMs in fact -- and in matches. That counts for a lot in my book. >>Of course while having this strength available it is important to be able to >>have a configurable opponent to different strengths and that is where HIARCS >>scores big time, with loads of progressive ELO levels which actually seem to >>make human like moves and mistakes at the level set! Not to just play at 2400 >>ELO one minute then drop to 1000 ELO for a move to drop a piece then back to >>2400 ELO seen in some other programs. > >you didn't get the point. I do not doubt that you are interested in a super >strong engine, however market research (done by Ubisoft and others) showed >clearly that for the huge majority of users, the playing strength is of no >importance. people who are interested in strong chess engines make about 0,00001 >percent of the population. >Besides from that, no other engine got so much bonus and lenghty discussions in >the review about it's playing strewnght than Hiarcs. It got the highest possible >score. It even won the prize for the strongest program. I even mentioned the GM >plays and furthermore even showed some games were Hiarcs played against GMs. You >should be the last complaining. > > > >> >>>> >>>>5) In the "Fischer Time" ranking, Palm Hiarcs scored 3/5 - Where is it missing >>>>functionality? You can even set fischer time on primary, secondary and the >>>>third time control. >>> >>>I can't comment on that in the moment. I don't own a copy of Hiarcs(Fischer time >>>is not usable in the demo). I tested Hiarcs on a friends device. But I think I >>>had reasons :-). If I will get a copy I might look at it more closely. >>> >> >>I thought you probably did not have an activated copy of the program from the >>inaccuracies in your review -- they seem to apply to the demo (free) version but >>not the full version. >> >>I think to review a program properly you need time to play with it and use the >>full version for some time or you are misrepresenting its capabilities and under >>valuing it. > >Agreed. O.K, then please send me a test version and I will be glad to correct >all incorrectnesses. I want to try to be as objective as possible, honestly. The >other companies let me test the programs (with the exception of Rampart chess). >Since you are directly involved in the development, you should profit from it. > >> >>>> >>>>6) There is an opening book style selector on Hiarcs but it scores 0. >>> >>>yes, I mixed that with the book selector, so the points from that column belong >>>in 'Adj. Opening book style'. Doesn't change the overall point though. >>> >> >>I see. >> >>>> >>>>7) Palm HIARCS scored 0 for "Tutor/Trainer" even though it clearly shows book >>>>moves, legal moves and incorporates a Coach Blunder Alert. >>> >>>This is all not covered under Tutor. For these attributes there are separate >>>categories. Hiarcs doesn't have a Tutor functions, such as Chessmaster has. >>> >> >>Did you not think the "find mistake", or "mistake analysis" or "displayable book >>moves with symbols" or any of the above warranted more than zero points? > >No. Again, there are already covered under different parameters. >That is the reason I made 100+ paramters, so that we don't have to sqeeze many >different thing into one number. > > >> >>Palm HIARCS is even able to point out a mistake in its own play and games which >>is a useful aid in understanding a game and what happened. >> >>>> >>>>8) Highlight Legal Moves scores 0 - This is odd because Palm HIARCS has a >>>>function called "highlightlegalmoves" which, suprise suprise, highlights legal >>>>moves! >>> >>>Agreed. That slipped the test. I will correct that in the next version. >>> >>>> >>>>9) "Move Entering" scores only 1, yet Palm HIARCS has normal move entry >>>>(from/to), smart entry (to) and 5 way key entry. >>> >>>It can't slide the pieces, which is the most important entry method. >>> >> >>I see, so you did not reward smart single tap move entry. How did you decide >>that "sliding the pieces" was the "most important entry method."? > >First of all, it's by far the most convenient. Secondly, it is the most >used/wanted way of move entry for most users (I have to add that I made a small >survey regarding used and wanted chess program parameters which included asking >several users and by getting survey results from other chess program >developers). 3rd, it is the most natural way of move entering, since it mimics >the way we do it in real live. > >> >>>> >>>>10) Playout scores 0 even though Palm HIARCS has autoplay support!? >>> >>>Yes, it has autoplayer (it got the full points for that !), but starting the >>>autoplayer may be not possible from a given board position (as mentioned before, >>>I can't test that anymore). If it can start autoplay after a board setup or from >>>within a game, then I can add points for that. >>> >> >>Yes, it can start auto play from any position or within a game. Then at the end >>of the game it can show you where mistakes were made allowing you to scan for >>mistakes forwards or backwards. > >O.K, will be corrected if true (I have to verify, of course how it works). > >> >>>> >>>>11) Save energy scores 0 but the set Elo levels ALL save energy. >>> >>>Oh, come on. Not playing the game will also save energy. >>>Other programs have special functions to lower the processor speed. That is >>>clearly missing in Hiarcs. >>> >> >>Hang on, it's you that has totally missed the point and usefulness of certain >>functions and abilities. The Set ELO levels uses very little CPU because HIARCS >>is so strong it thinks to the required level very fast so the battery IS saved. >> >>It also means that on other levels with HIARCS I can allow it to think for less >>time and get the same quality as other Palm chess programs thinking for much >>longer. This obviously saves energy because the CPU is used for less time! > >That statement is not supported by any data and makes not much sense in terms of >usability. Also, this is true for all programs, so no harm for Hiarcs. Don't >come with strength, that is covered elsewhere. And again, as I said before, if >you don't use the program, you save even more energy. The point is, that you may >need to decrease the processor usage in some cases, when needed for other >things. It is most convenient if you have a function for that. But I can make a >compromise; I will lower the points awarded for this function from 2 points to 1 >point. > >> >>Maybe you should suggest to Mark that he puts a save energy checkbox on the next >>version so it is clear that energy is being saved! > >Would actually be useful, since most users will have no idea about it. The way >programs decrease their strength do not have necessarily anything to do with the >processor. How is Hiarcs handling it? Usually, the strenght is handicapped by >changing engine parameters, leading to inaccurate play. No energy is saved that >way. If you switch to a shorter time control, the players time is also lowered! >So that is no option. > >> >>>> >>>>12) Next best scores only 1/2 - However, Palm HIARCS has a full next best >>>>function -- certainly more than Palm Chess Genius which scored 2/2 but does not >>>>have the function at all!?!? >>> >>>You are incorrect. Chess Genius HAS this function and it is easier accesible >>>than in Hiarcs (one step less to reach the function; which is important if you >>>play, you don't want to wander through menus). >>> >> >>Do you even own a Palm? >> >>I am running Palm Chess Genius 2.22 (latest version) and it does not have a menu >>item or option for "next best." >> >>I am beginning to think you have not done a proper job at all here, you have >>reward Palm Chess Genius for a function IT DOES NOT HAVE and you have not >>properly rewarded Palm HIARCS for a function IT DOES HAVE! >> >>By the way if this one tap functionality is so important maybe you should give >>points for Palm HIARCS one touch move now, analysis and start thinking features. > >I did. > >> >>>> >>>>My next question would be: Did you actually test and run Palm HIARCS yourself >>>>for this review? >>> >>>Yes, I did. Unfortunately, as I said, I don't have it available anymore. Maybe >>>Mark Uniacke can give me a test version, I don't know. >>>Most points you raised I could explain as you could see. Then, I made one >>>mistake, which is not surprising (and I expected that when testing several >>>thousand parameters). But for that we have people like you who can find these >>>mistakes. Thanks. >>> >> >>What you did explain gives the strong indication to me that you have not >>properly reviewed Palm HIARCS and maybe other Palm programs (you said Palm >>Genius had next best etc.) At best you gave Palm HIARCS a cursory look over, at >>worst you only used the demo version without full features or chess strength! > >I tested Hiarcs as best as possible. Yet, I admit I tested it the shortest time >of all programs, since it is the only program (besides Rampart chess) where I >don't have the program and where i had to test on another persons Palm (keeping >in mind that due my fairness and responsibilty, I didn't even asked for an >illegal copy of him!). It would be better for users if you guys would offer a >fully functional demo version (or at least a test version for reviewers), so >that the proram can be tested accordingly. Of course, I don't want to introduce >mistakes in a test. I like Hiarcs and will be very glad if I can improve the >review in regard to Hiarcs. > >Best >Axel Axel, We could hash out the above points for many more messages and you still did not address the issue of the Chess Genius v2.2 "next best" feature that you say is present but I cannot verify. Rather than continue to ride a dead horse, I'll leave things with this note: I believe I've brought forth many valid points that ARE important to the average Palm/Pocket PC Chess Software connoisseur and continuing to hammer on them will not help or change anything if you are determined to have a subjective-based and rushed review over a logical, well-thought out and factual-based review. Specific to HIARCS, you have presented a review as if you owned or had a review copy of Palm HIARCS when in fact you only had a free restricted feature version or very limited access to another copy. As a result, you have made many mistakes in the review of Palm HIARCS and have misrepresented it. You also made claims for another program which were wrong and you even compounded the error by insisting in this forum that you were right when you clearly had not checked out your facts. You admit you tested Palm HIARCS the "shortest time of all programs" yet you say you "tested Hiarcs as best as possible"? I don't see how those two statements line up. I would have thought that given the fact that Palm HIARCS is a milestone in Palm chess software, it deserved more than a brief test with a restricted free demo version! You complain that you were not sent a review copy, but did you ever ask for one? It would seem very unfair to complain that Palm HIARCS was the only program apart from Rampart that you did not receive a review copy for when you never requested it!? To the VAT issue: Lang Software may not have to pay VAT if they registered as a small business exempt from VAT tax so you do not know if they pay VAT (reducing their revenue) or not. It is you that is talking nonsense when you say the "best way" is to present a price which is not the price the customer pays in most of the world. VAT is a regional tax and is added in the respective regions if necessary. For example, it is European Union tax law that VAT should be charged to purchasers residing in the European Union -- outside of the European Union there is no tax charged (in fact VAT actually varies depending on the European Union origin & destination country!) Hence, it makes no sense to include VAT in each price as most of the people do not have to pay it! When a person goes to the secure server to purchase Palm HIARCS, they immediately see on the first screen the exact price they need to pay before they proceed, please don't continue to misrepresent what happens. If I go to the Palm HIARCS order page I get the price $38.29. That is the final price I pay. I think it is more sensible and straight forward to make a review with the non tax price and make a note that sales tax may be applicable in some regions. Don't assume everyone lives in the European Union! Regards, Enrico L. Carrisco
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.