Reply to thread

Message

<blockquote data-quote="LightPhoenix" data-source="post: 5378552" data-attributes="member: 115">I don't mean to impugn your honesty; I'm sure that's what you were told.&nbsp; However, I highly doubt this is something that could not have been caught before release.&nbsp; In house stress-testing is good, and nothing is loading slow for me, so obviously the conclusions based on it were correct.&nbsp; However, any web-based application is going to need &quot;real setup&quot; (as opposed to in house or approximate setup) testing, and it appears that is was not done.&nbsp; The reason I say that is because the crashes are so prevalent that I find it hard to believe they would not have been caught if that had been done.I don't think it's due to server load either, at least not load on the program.&nbsp; Based on my experiences with the crashes, my (entirely ignorant) guess would be some sort of server hiccup in communication between the program and the database(s).&nbsp; That could account for the randomness of it, and it always seems to happen when choosing an option.&nbsp; It could be due to load stress on the database, and not the website.&nbsp; I'm not certain exactly how the stress testing was done, so I don't know if that's the cause.If I may forward a suggestion on to the developers - this seems like something that could be easily avoided if some sort of error handling was in place.&nbsp; Perhaps if an error occurs, the program simply doesn't add the feature, and displays a message apologizing and asking the user to pick again.</blockquote>

[QUOTE="LightPhoenix, post: 5378552, member: 115"] I don't mean to impugn your honesty; I'm sure that's what you were told. However, I highly doubt this is something that could not have been caught before release. In house stress-testing is good, and nothing is loading slow for me, so obviously the conclusions based on it were correct. However, any web-based application is going to need "real setup" (as opposed to in house or approximate setup) testing, and it appears that is was not done. The reason I say that is because the crashes are so prevalent that I find it hard to believe they would not have been caught if that had been done. I don't think it's due to server load either, at least not load on the program. Based on my experiences with the crashes, my (entirely ignorant) guess would be some sort of server hiccup in communication between the program and the database(s). That could account for the randomness of it, and it always seems to happen when choosing an option. It could be due to load stress on the [I]database[/I], and not the website. I'm not certain exactly how the stress testing was done, so I don't know if that's the cause. If I may forward a suggestion on to the developers - this seems like something that could be easily avoided if some sort of error handling was in place. Perhaps if an error occurs, the program simply doesn't add the feature, and displays a message apologizing and asking the user to pick again. [/QUOTE]

Verification