Hi Beta Testers,
We are currently stumped with a memory issue with Windows. Interestingly, we can not reproduce the bug under Linux as frequently. The bottom line is we are going to need more time to fix this issue and utilize some Memory Error Detection tools to assist us. We are pretty sure the bug has to do with creating lists since we made changes to enable nested lists as well. This seems to have a side effect in the rtf writer; however, this may not be the case. As you’ll note in the screen shot below, a nested list - this took six crashes under Windows to complete, under Linux zero? I created the same list yesterday under Windows without a single crash?
So, what does that mean for you. Well, we’re in the process of compiling a full Linux beta 1.6 build which I will upload as soon as it’s done and then edit this post with a download link. This has the full code. We are hoping, given that the bug seems less temperamental under Linux, that releasing this to geeky Linux types might assist us gain some insight into identifying and reliably reproducing this bug by following exact steps they might propose?
For Windows, we are stripping out the nested list changes and the fix that prevented loosing first line indenting and formatting - as the two are tied. What this should give us is a reasonably stable Windows build which we’ll call Beta 1.56 (I know it’s getting messy).
The upside of this approach is that all platforms will get many fixes and enhancements. Probably the ones that are likely to bring the most joy are dictionary fixes to learn and remember words, unlearn words, download foreign dictionaries, and the ability to customize the main toolbar and also the formatting tool bar so you can set up Scrivener however takes your fancy. The toolbar customization functionality is not drag and drop enabled currently, but it works well and allows complete control. See screen shot below:
Windows users who have MS Word installed will also be able to take advantage of the verydoc conversion tool within Scrivener and export their work as doc, docx, html, odt, many image formats, etc. Import using the same engine needs to wired up and has been put on the shelf until we can resolve these memory bug issues.
So, this is not how I imagined this release going - Murphy’s law proved a certainty.
I will post more once the builds are uploaded.
1st Feb 2011, 10:15am +10EST