Chris

Hello Walt, Thank you for your feedback! The "Index content embedded in documents" option was indeed ignored in 1.8 beta 1 (= acted as if the user had it turned on). Beta 2 is about to be released and will fix this. Adding more fine-grained permissions for Connect is on the roadmap. I have no explanations for the issues with the Type and Location facets. Perhaps you can open a support ticket and share the case logs with us?

Actually it's been in Intella for quite some time already. Just right-click on any of the results and select "Export queries".

Hello Gary, Thank you for your feedback! The Intella EnScript is a contribution from our user community. I will forward this information to the creator of the EnScript. As for the Location facet, I will ask our developers what we can do to make this more usable.

Interesting idea! Note that this will only work for emails when you index with the same Intella version. MD5's will be the same but message hashes depend on some decoded fields and MIME message data structures. Any improvements in message processing, e.g. to better respond to broken data structures in a PST file, can then affect these hashes. A CSV for each tag with MD5, Message Hash and Message-ID columns would be a good approach IMO.

Many thanks for sharing! The 1.8 beta indeed seems to ignore the embedded items setting, it simply always extracts all embedded items. We will make sure that gets solved in the next beta. Note that that will make the total indexing duration with 1.8 even shorter, as now it is unnecessarily doing more than 1.7.3 did!

When this keeps happening with beta 1, could you share the logs with us via a support ticket?

Hello Philip, An email data set usually contains a fair amount of forwarded messages, added as an attachment to their parent mail. These forwarded messages are counted as separate items in the Intella case. When you export all items to a PST with the default PST export settings, the forwarded mail will be replaced by its parent. As the parent is already in the set of items to export, it will not be added twice. Therefore Intella will report two exported emails, whereas Outlook at first glance only shows you one - you need to open the parent mail to see the attached forwarded mail. The same applies for all items nested in any attachments, which is why the unique item count can still be a lot larger than the number of messages visible in Outlook.

When you try to export a file to a PST, Intella will export the parent mail/meeting request/contact/... instead, unless you modify the default export settings. See the "PST options" section of the user manual for more information. Any items that cannot be exported are always listed at the end of the export report, together with an error message.

Hello Adam, That is correct. At the moment the 1.8 beta can only open cases made with it, 1.7.3 cases will be listed as disabled. This is due to the many database changes that were necessary for the improved indexing performance. The final version will have that functionality.

Hello Philip, Please send us the logs listed above, only that way we can effectively help you.

I would suggest that you create a support ticket and share more specific info for which this forum is not the right place. Then we can look into what is causing the crashes on your system. Information we would be interested in: The case log files of the case that crashed. All hs_err_pidXXX.log files in your temp folder (type %TEMP% in Windows Explorer), where XXX is an arbitrary number.

Hello all, I am proud to announce the first beta of Intella Connect 1.8, featuring the following enhancements: Improved general design and layout, including a more appealing login page. Connect can now be installed as a Windows service. Connect now runs on a single port. Previously the admin page and each shared case needed its own port. Single sign-on lets reviewers access all their cases without having to re-enter their credentials for each case. Added a case dashboard, giving an instant overview of key case characteristics and events. Relaxed the rules on what legal user names are, and now allowing users to change their own password. Extended permissions model with the ability to prohibit printing and to allow deletion of tags made by other reviewers. Added case auto-discovery: all cases stored in a folder configured in the Connect server's settings are automatically available for sharing in the admin page. Added the ability to manually add a case to the server by specifying its case.xml file. The Contents tab now captures the original layout better, e.g. tables, font sizes and font styles are shown. Added various table columns that were previously only available in the Desktop edition. Added support for viewing images not supported by web browsers, such as BMP and TIFF, by converting them to PNG on the server. Added Content Analysis facet. Various fixes to make Connect run properly on Internet Explorer 8 and 9. Forum users who would like to test the software are invited to reply to this topic or send me a private message. I will reply with a private message with instructions. We look forward to your feedback! For example, we would like to hear how the new user interface is working for you and whether the above improvements make deploying Connect in your infrastructure easier.

Hello all, I am proud to announce the first beta of Intella 1.8. This version features many enhancements. A list of the most prominent ones: Greatly improved indexing speed; up to three times faster indexing time, depending on data set and hardware. Paragraph features: collapse/expand paragraphs, search for selected paragraphs in other items, optionally hide already seen paragraphs. Task wizard, for defining common processing steps such as applying a keyword list and tagging or exporting the results. These can be run automatically once indexing has completed or by manually starting them. Improved disk image processing: faster opening and processing, added support for MacOS, GPT and ISO images. The new Sets display shows how search results compare in size, without showing the overlaps between the results. The Contents tab now captures the original layout better, e.g. tables, font sizes and font styles are shown. Redaction refinements: added Delete button and Undo/Redo support. Prevent complex Cluster Maps from freezing the application. Added support for Bloomberg email dumps. A critical fix for Intella's dongle driver, necessary for users running on Windows 8.1. Forum users who would like to test the new release are invited to reply to this topic or send me a private message. I will reply with a private message with instructions. We look forward to your feedback! In particular we would like to hear about indexing performance stats: what type of size of data set did you index, how long did it take, what did you see in the Windows Resource Monitor, etc.

Hi Jason, Thanks for sharing! Would you agree that the procedure of setting the "Is Attachment" marker only marks the actually attached files, or does it also need to be set on items embedded in those attachments? E.g. images in a Word documents, files inside a ZIP file, etc. Note that you can get to these additional items using the Show Children method, so this is only about convenience.

One thing I'd like to add: since Windows Vista, Windows got a lot better in putting the free memory to good use, basically letting it work like a sort of disk cache. This means that the negative impact of using a slow drive such as a WD Green disk could be alleviated or even entirely masked by the data caching in the free RAM, as long as you don't have a peak in application memory usage and the disk data that is being manipulated fits inside the free memory. I recently upgraded an older test machine from 8 to 16 GB of RAM. The time needed for an indexing test that I routinely perform on it (77 GBs of PSTs) went from 9 to 6 hours. The impact is probably higher on larger cases due to the random access on the database files in the case folder: random access isn't an issue for RAM but it is for spinning disks.

I did, and I must admit that it still has not been implemented I will give it a boost in our feature tracking system.

I will ask our developers if we can add columns and Features facet categories for identifying attachments (given to all items whose direct parent is an email, MMS message, ...) and embedded items (given to all items where one of the parents is a document, spreadsheet or presentation). Other categories that are already planned are Recovered items and Orphan items. If you know of any other item categories that would make your life much easier, do let us know!

Regarding indexing post-processing tasks, incl. exporting, see my posting here: http://community.vound-software.com/index.php?/topic/171-wish-list-intella-scripting-engine/?p=845

Hello Phil and Adam, Combining cases is certainly on the roadmap! At the moment we are close to releasing a beta of the next major Intella version, which focuses on indexing speed and large case handling, among other things. The ability to merge cases is not present yet, but the architectural changes provide a much better basis for adding often-requested features such as merging cases, exporting items as a separate case, deleting items from a case, etc.

Hello Mark, In your other post (http://community.vound-software.com/index.php?/topic/171-wish-list-intella-scripting-engine/) I already mentioned the post-processing tasks that we will add in the next version, but I am not sure if that covers what you need here. Do you want to apply the tags based on keywords, or is this merely about populating the list of tags available in the Add Tags dialog?

Hello Gary, Your message puzzled me a bit as there is no "_to long paths" or "_too long paths" in Intella's source code, but I believe I have figured it out. Could it be that you are using the Intella EnScript to export items from a disk image to a folder? This script uses this folder (incl. the typo) for paths of a certain length. If I recall correctly, it applies rather strict rules on maximum path and file name lengths so that the exported files can still be manipulated with Windows Explorer, which has much stricter rules than the NTFS file system that it is usually working on. Intella has no issues indexing and exporting the much longer path names that NTFS allows. If the issue is indeed with the EnScript, please ensure that you are using the latest versions of the script (v0.5) and Intella: the latest script can export to logical evidence files, which can be indexed by Intella. This will be a lot faster and keeps the full path intact.

Hello Mark, I am happy to tell you that in the next version you will be able to define index post-processing tasks, i.e. find all items matching a keyword, keyword list, MD5 list, date range or an arbitrarily complex Saved Search and let them be tagged, flagged, commented or exported using an export template. It is configured through a UI and therefore may not be as flexible as a full-fledged scripting language, but it will surely make a vast difference.

Hi Adam, Are you able to share this email with us? A corrupt email is the most likely explanation as else we would be missing a lot of data, but I would like to rule out any processing bugs. Ideally, a copy of the mail copied to a new PST file works best, as then all PST-specific fields are kept as well.

There is already an alternative method. Note the second-to-last button in the Cluster Map's toolbar. When selected, only the most-connected clusters are shown (with a maximum of seven, if I recall correctly). You can combine this with a rank-based color scheme (see the third button in the toolbar), which colors clusters based on how many search result sets they relate to. See the attached image. Here you can immediately see that there is only one cluster that links to all five search terms. All other clusters always lack one search term. I had to redrag some clusters to improve the graph layout.

The way it works now is that it displays the union of all selected sets, i.e. as if they were merged into a single set. However, I can see the use for displaying the intersection instead. I'll give it some thought...