Results 1 to 6 of 6
  1. #1
    Join Date
    Dec 2012
    Location
    Huntsville, AL
    Posts
    379
    Provided Answers: 2

    Answered: Inconsistent Results From Scripting

    I have been hesitant because of the expense of the program but I was finally ALMOST poised to actually purchase the Ultimate version of Brilliant Database as I love the idea of being able to make stand-alone versions of a program! Granted, I, unlike others here, have dealt with some very questionable results from the Brilliant Database product (i.e., queries that did not yield the proper results as well as other outputs that just did not seem right… well, if the truth be told, they were not right). Please know that these were not one-time occurrences but sporadic results which actually troubles me much more.

    My latest experience was over this past weekend and I ran a report that took several hours to complete. Upon seeing the results (which appeared to be accurate and correct), I was forced to rerun the identical report because of the known bug with naming conventions and optimizing (yes, I have learned to save the program prior to doing anything ‘major’ in the event that I have to start over). When I saw the new results, I knew immediately that they were vastly different… and incorrect – this was using the same, original file with ABSOLUTELY no changes. I ran the lengthy report again and got incorrect results yet a third time. I closed the file, re-opened the earlier 'good' file and ran it a fourth time this morning only to get totally different results. I have seen too many instances with Brilliant Database being unstable, inconsistent, and providing invalid results.

    Although everyone programs differently, my scripting is sound and valid though I repeatedly see that the results are not always repeatable. This troubles me greatly as my project this deals with physicians credentialing and it MUST be accurate. I know that I can create a program in EXCEL with Visual Basic for Applications that runs considerably quicker and yields consistent and proper results but I do not want to start at the beginning simply because of the major shortcomings of Brilliant Database.

    This program may be ideal for capturing some basic readings; however, it is my experience that it is not suited for doing any sort of complex analysis. Couple this shortcoming to the cost of the program and TOTAL LACK of customer support and it is not a wise investment for me. It is troubling that in the rare event that you actually get a reply when you send a notification of a bug to Brilliant Database, and it is acknowledged as a bug, and yet the bug continues to exist - this is not good for any organization. Some may state that they have not experienced any issues but I have had many!... too many!!!

    I am in a terrible bind here as I have been contacted for using this program again and, because of the critical nature of these certifications, I have no confidence in the validity of the results as they appear to change from run to run with NO changes in data in Brilliant Database. I do not have the time to start anew and recreate a workable program in EXCEL. I am troubled because the physicians training program has no backup plan – just a lot of data with no viable program to analyze the information accurately. If anyone has any suggestions for a programmable database that actually yields accurate information and is affordable, please let me know because I have reached my utmost frustration level with this program.

    I would love to be an advocate for Brilliant Database but I cannot because it simply does not always work. Easy stuff (i.e., grabbing data and sorting) it appears to wok fine but with more complex analysis it fails miserably. I am quite bitter towards the Brilliant Database product right now...
    Daryl G

  2. Best Answer
    Posted by Tubbritt

    "Hi Daryl.
    Just doing some experiments here and I found something interesting.

    The Query function seems to try to push every single record it finds into it's memory, even if it's specifically told to only process the first 1000 records. This is overloading it and causing instability to the point where I can crash brilliant database every single time.

    It shouldn't be doing that.

    For example, the normal “find” option will abort by default any search that finds over 1000 records from a folder that contains 1 million records. Where a query will fail 100% on 1 million records regardless of the limit applied to it's function.

    Sounds like something brilliant database could fix very easily if they could be made aware of this. Simply put, the query function should not be trying to remember every record if it's told to only remember 1000.

    Regards
    James."


  3. #2
    Join Date
    Dec 2012
    Location
    Huntsville, AL
    Posts
    379
    Provided Answers: 2
    I would like to clarify this with a little bit of additional data. The scripting in question consists of just 25 lines of code - seven of those 25 are comments or messages. The scripting is not complicated consisting of one outer 'If' statement, two nested 'If' statements with two 'For Each Record... Next Item' command in between. There are no issues with the scripting; however, Brilliant Database provides different results when running the script there are NO changes to any of the records. If I later discover that I have mistakenly messed up and the problem is with something I have done, I will immediately apologize here for my accusations (I have checked and double-checked and the script is sound). I am troubled by these results as my trust in this program are lacking. Too many times have I seen some highly questionable behavior from this program.

    On a positive note: the dBforums Brilliant Database area is wonderful and I appreciate all the help that was given to me.
    Daryl G

  4. #3
    Join Date
    Jul 2011
    Posts
    176
    Quote Originally Posted by fireant911 View Post
    I would like to clarify this with a little bit of additional data. The scripting in question consists of just 25 lines of code - seven of those 25 are comments or messages. The scripting is not complicated consisting of one outer 'If' statement, two nested 'If' statements with two 'For Each Record... Next Item' command in between. There are no issues with the scripting; however, Brilliant Database provides different results when running the script there are NO changes to any of the records. If I later discover that I have mistakenly messed up and the problem is with something I have done, I will immediately apologize here for my accusations (I have checked and double-checked and the script is sound). I am troubled by these results as my trust in this program are lacking. Too many times have I seen some highly questionable behavior from this program.

    On a positive note: the dBforums Brilliant Database area is wonderful and I appreciate all the help that was given to me.
    In the past I have created reports using over 500 lines of script but never faced such issues except slow in response(very slow sometimes 10 minutes of waiting).

  5. #4
    Join Date
    Dec 2012
    Location
    Huntsville, AL
    Posts
    379
    Provided Answers: 2
    asokram,
    This scripting in question is going through 100,000 records and looking for certain records in multiple folders to move to an Archive Folder. The original idea was to that I wanted to delete those records but there is a bug with Brilliant Database that causes problems when deleting records via script. So, I now send the records to the Archive Folder, manually delete the records from the Archive Folder, and then finally delete the records from the Recycle Bin. The issue is that the records the program is looking for are depends upon a check mark being present in a flag field... which is really straightforward. Running the program takes several hours and the end result often contain records that do not contain a check mark in the flag field; however, sometimes, just the proper records are sent to the Archive Folder (just as it is supposed to do). Most troubling is that re-opening an identical version of the program and running it again often yields different records... but not always! I do not know if Brilliant Database is having issues with the volume of records (since it is at the maximum number of records in a folder - at least according to the BD website).

    I know that every program has some quirky behaviors that developers have to learn to live with but some of these issues go well beyond quirks.
    Daryl G

  6. #5
    Join Date
    Jul 2012
    Location
    Ireland
    Posts
    815
    Provided Answers: 17
    Hi Daryl.
    Just doing some experiments here and I found something interesting.

    The Query function seems to try to push every single record it finds into it's memory, even if it's specifically told to only process the first 1000 records. This is overloading it and causing instability to the point where I can crash brilliant database every single time.

    It shouldn't be doing that.

    For example, the normal “find” option will abort by default any search that finds over 1000 records from a folder that contains 1 million records. Where a query will fail 100% on 1 million records regardless of the limit applied to it's function.

    Sounds like something brilliant database could fix very easily if they could be made aware of this. Simply put, the query function should not be trying to remember every record if it's told to only remember 1000.

    Regards
    James.

  7. #6
    Join Date
    Dec 2012
    Location
    Huntsville, AL
    Posts
    379
    Provided Answers: 2
    James,
    Thanks for looking into that and your observations are very much in line with mine. I know that I seem to be having more issues than most people do but, after reading your reply, I do not feel as crazy as I did before (but that does not negate that I have more problems than everyone else does). I think the folks here have done a tremendous amount of beta testing of this program and if these issues were corrected, this would be an INCREDIBLE program. The lackadaisical response from Brilliant Database has been brought up repeatedly and there is no value in saying it again. This program has a HUGE potential but until Brilliant Database listens to the voice of their customer, they are severely limiting the marketability of this product.

    I have made a work around that appears to do okay although it requires more user interaction that the earlier approach (the scripting is identical - the only difference is that only one doctor may be selected at a time instead of multiple doctors). The volume, as you alluded to, may be overwhelming the program.

    I still want to thank all the great people here for their continued support, help, patience, and guidance!
    Daryl G

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •