A couple of minor niggles

[roq]

Richard

Hi.

Just a couple of small things - On firefox:

- Sometimes the problems slow to a crawl and eventually after the success flag appears the problem freezes and the result is not credited (this invariably happens when I've just solved a really hard one in no time at all of course  ). Note that the internet connection
remains up and I can navigate to other areas of the site.

- Sometimes in the problem screen, the system automatically turns on the "continue if correct option" (which I never use) by putting a check in the box.
 
Haven't had either of these problems on IE yet, but don't use IE much.

Neither of these minor problems are particularly earth shattering, but just thought I'd let u know.

[richard]

Thanks Roq,

On the "continue if correct option" problem, have you been noticing this for a while or is it something that has cropped up recently? (I'm worried it might be a regression from my recent "fix" for problem type selection).

I assume from your mention of IE that you are using Firefox on windows.  I haven't seen that issue under Mac OS X yet. I only use Windows briefly to make sure UI changes don't fundamentally break that platform so I sometimes don't catch intermittent problems such as the one you describe.  Have you noticed any pattern on when the problem occurs? (other than when you've solved a hard problem in 2 seconds :-)  ).  Also which version of Firefox are you running?

Thanks for the problem reports.

Regards,
Richard.

[roq]

On the "continue if correct option" problem, have you been noticing this for a while or is it something that has cropped up recently? (I'm worried it might be a regression from my recent "fix" for problem type selection).

This is something I've noticed for some time so it is unlikely to be the result of a recent fix. I'll try to pin down more precisely the circumstances in which it occurs - I know that this kind of problem can be very hard to identify, particularly considering the diverse behaviour of different browsers on different platforms.   

I assume from your mention of IE that you are using Firefox on windows.  I haven't seen that issue under Mac OS X yet. I only use Windows briefly to make sure UI changes don't fundamentally break that platform so I sometimes don't catch intermittent problems such as the one you describe.  Have you noticed any pattern on when the problem occurs? (other than when you've solved a hard problem in 2 seconds :-)  ).  Also which version of Firefox are you running?

No I haven't really detected any pattern, but it is progressive - problem solving slows down and finally it hangs on a problem with the clock still ticking. Interestingly i sometimes also get overticking whereby the clock continues ticking after a problem is done - so it maybe a routing delay problem (I assume the ticking code is local). I don't see this as something that needs an urgent fix, rather something to bear in mind if you find that others are experiencing this kind of problem.

Firefox is latest version I think:

Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8.1.11) Gecko/20071127 Firefox/2.0.0.11

[richard]

Thanks.  It sounds like the slowing down problem is most likely a memory leak.  I rarely do more than 50 or so problems at a time which is probably why I don't notice it.  Does the rest of the browser become sluggish or is it only the ChessTempo window/tab?

Thanks,
Richard.

[roq]

Thanks.  It sounds like the slowing down problem is most likely a memory leak.  I rarely do more than 50 or so problems at a time which is probably why I don't notice it.  Does the rest of the browser become sluggish or is it only the ChessTempo window/tab?

Thanks,
Richard.

Yes now you mention it does behave as if it might be memory leak or resource problem - Only the CT window slows down and the rest of the browser is unaffected.

[richard]

Yes now you mention it does behave as if it might be memory leak or resource problem - Only the CT window slows down and the rest of the browser is unaffected.

Ok thanks, I'll try testing with some longer sessions.  I also barely test on windows under firefox, most of the testing I do there is for IE, so that may also be a factor.  If you could let me know how many problems into a session its start to get sluggish that would also be helpful.

Thanks,
Richard.

[roq]

If you could let me know how many problems into a session its start to get sluggish that would also be helpful.

Hi again

In my last session did about 70 problems. At around 45 problems it became slow to move to new problems and the report of score slowed down (but not on every problem). Sometimes also the first move was slowed down. At just over 60 problems, I had an overticking event where the clock keeps ticking after the problem has been solved (and credited). Approaching 70 the slow down and movement between problems appeared to become more marked... Didn't fail in this session though.
 

[richard]

Ok thanks, that gives me enough info to go on with.  Hopefully its something I'm doing wrong and not a browser issues which requires crossing of fingers and waiting for the Firefox guys to fix it :-/

Regards,
Richard.

[roq]

Hi Again

I've been trying IE v7.0.5730.11 some more and have now experienced both the glitches - so it is not restricted to Firefox. Also the glitches don't seem to be arising consistently after a set number of problems as I had speculated earlier and I can't tie down just exactly in what circumstances they occur. One thing though - I often switch to look at the stats page (to check accuracy) so that may reset something?

Of particular note is that the credit display is slow and sometimes the clock ticks for several seconds after a problem has been marked correct, but before the score is credited.     

[richard]

Ok, I'm glad its happening on both platforms as it is more likely that the problem is under my control.  I probably will not have time to look at it under after Christmas but hopefully it will not take too long to track down.

Regards,
Richard.

[roq]

An update - I haven't had the first problem recently at all. However, "the continue if correct option" is still being altered. I think this happens when one moves away from solving tactics to another screen and then click solve tactics again. For instance if I go to the stats screen then click on a problem number go the problem screen and then click solve tactics the system automatically puts a tick mark in the "continue if correct" box.

P.S. I think this option should be unticked by default as that might encourage more people to contribute ratings and comments.   

[richard]

The encouragement point is a good one.  Thanks for the info on when the problem occurs, that should be enough for me to be able to hunt it down.

Regards,
RIchard.

[roq]

I think I know what may be happening - The problem view screen and the Solve tactics screen are variations on the same basic screen. I suspect that when the screen loads the flags are set to their default values, and perhaps the problem view screen does not check whether the user has altered the flag setting whilst the solve screen does.

You could just change the flag default to off rather than on - then it wouldn't be a bug but a feature! The system automagically turns the flag off when you review a problem so that you can now review problems in situ! What a helpful UI - almost as good as MS's talking paperclip :-)

[richard]

Hi Roq,

I think you're assessment is bang on.  Should be fairly easy to fix.

Thanks,
Richard.

[macleay]

I have had this problem too. Problem continues on correct when I don't select it. Also changes from standard to blitz unawares.