Q&A Dec 2017

This is an edited collection of questions and answers following on from the presentation was given to RUG (Roots Users Group) in December 2017.

Q: When is HRE likely to be available?

▪ A: This is highly dependent on our ability to attract coding resources. Our last detailed analysis suggested around 550 person-days of effort.   We are deliberately not promising to release the product by a specific date.

Q: How many volunteers do you actually have who are currently working on coding?

▪ A: The ‘Core Team’ currently includes 4 members with a variety of coding and database skills.

Q: Given that you have only a small coding community so far, how will you attract enough additional coders to finish the task?

▪ A: Pro bono work from employees of large history or computing organizations or teaming with an education authority to use HRE development as a student exercise are pathways we have not yet explored.

Q: If RUG (or its members) were able to offer you significant additional funding, would you want it? What would you do with it? How if at all would it help?

▪ A: Given the potential 550-day effort, if a person-day cost only $500, then the total budget required would be $275,000. We do not regard this as a viable path, nor do we want HRE Ltd handling the legal obligations of employees.  We may at a later stage want to seek funding for the completion of specific program areas, but we are not at that point yet.

Q: How do I register to be a Beta tester?

▪ A: Click here to join our mailing list. When we are ready to accept beta testers we will announce that on the list.

▪ Q: How can I make a donation?

▪ A: Click here, or here to discuss corporate sponsorship.

Q: I’m not technical and can’t help with coding. Is there anything else I can do to help? What would you like me to do?

▪ A: We have a large pool of people who have put up their hand for testing and documentation assistance, but more assistance is needed, especially with detailed coding. Offering to assist in the development of HRE must be seen by the volunteer as a commitment. It will require dedication of time over a longer period. Please consider your position before you volunteer to ensure that you understand the responsibility of becoming involved with the HRE project.

Q: What are some of the areas where you need help? If we have developers or IT people in our social networks who know nothing about TMG, would it be helpful if they volunteered?

▪ A: Yes, absolutely, provide they have skills in a relevant area such as working with an integrated development environment (eg Eclipse or Netbeans), or have skills in Java,  another object-oriented language, database management, or GUI design.  You can find out more about our current needs on this page, and register your interest here.

Q: Recent forum discussions have stated that HRE will have website creation  functionality like John Cardinal’s Second Site or GedSite. Is this true?

▪ A: No. The HRE roadmap does not  currently include the functionality of Second Site or GedSite. HRE development plans do include a GEDCOM export at pre-release v0.6, which a future GedSite version could potentially process. However as GEDCOM can only handle a small fraction of HRE’s functionality,  that approach will be unsatisfactory without very significant work.  John Cardinal’s Second Site program as currently written is intimately intertwined with TMG and has no immediate relevance to HRE.

More promisingly for the longer term, HRE will offer a full data export, probably in XML format. We would welcome the development by John or any third party of new HRE-specific standalone software, along the lines of Second Site, to create websites based on such HRE output.

Q: Can we assume that since the database structure will be open, presentation of the data could be accomplished by third parties, as with Second Site for TMG?

▪ A:  Yes, we plan to make HRE’s XML output as open as possible so that the data can be freely read by third-party software.

Q: Recent forum discussions have pointed out the size limitations of the TMG database, where if ANY table exceeds 2 Gigabytes, the database becomes corrupt. What is the limit on H2 databases?

▪ A: The largest allowed H2 file is 2 Terabytes, provided it is not resident on a device formatted in FAT or FAT32 mode. Or to put that another way, the H2 database limit is 1024 times the TMG table limit. The forum discussion concerned a TMG database containing slightly over 1 million names, which had pushed the TMG Name table to the 2GB limit. Without knowing the size of all the other TMG tables in this example it is not simple to compare with H2, but we can estimate that an H2 database should be able to handle several hundred million people or other entities.

Q: It all looks very complex and complicated. Will there be a beginner mode and perhaps a more advanced mode?

▪ A: We understand there are users across a range of experience and understanding. We expect to build a user interface that allows users to learn from a simple position and grow their knowledge out to include the features they wish to use. Potential users have identified the need for two interfaces – a full version for researchers and a simplified one just for people collecting data. We can either build a lot of detail into a single screen, or have a chain of screens where one screen opens another which opens another, etc. The challenge is balancing these designs so that it is easy for common things to be done with minimal mouse clicks/keystrokes without locking out the ability to implement more sophisticated features.

Q: Is HRE designed as a TMG replacement or will there be a version that is less sophisticated?

▪ A: What we think you’re really requesting is a simpler user interface. We are not going to change the internals of HRE to create a simpler version. What we have to do is hide ‘complexity’ in a better user interface so that you can choose what features you wish to see, and not be presented with those you have no interest in or perceive as too complex. We will make our own best guess, but will need input from users, who will no doubt have other views. Our aim is to create a  flexible solution.

Q: What functionality within TMG will NOT be in the first release?

▪ A: As shown in the presentation, there was a road-map for V0.1 – 0.6 pre-releases. We expect the last stage (reports) to be broken up into a number of progressively more complex reporting capabilities. So for some people, HRE may appear to be ‘complete’ earlier than those who may want to see ALL reports done before they accept HRE as  being ‘complete’.

Q: What interest  have you had from people interested in the history aspects of all this?

▪ A: Simple answer – a lot of interest. As an example, Robin spent time last week at a conference in Canberra where academic historians (some using TMG) had come together with family historians. The realisation was that these two groups have grown together in their needs for tools to gather and analyse data. We have also been in touch with other groups who are potential HRE users for the recording of the history of their organisations and their holdings of historical artefacts.

Q: What TMG functions do I LOSE by going to HRE?

▪ A: HRE will not support the concept of Data Sets within a Project, but WILL allow individual Data Sets to be imported as a Project. Soundex values will also not be stored in HRE. Some Timelines cannot be transferred from TMG into HRE, as their copyright is not held by Wholly Genes.

Q: What imports will be available?

▪ A: The only imports planned by the HRE team are from TMG projects of v8.05 or later, and GEDCOM. As the design of HRE allows for additional ‘plugin’ capability, others may add further import functionality.

Q: When will the imports be available?

▪ A: TMG import capability should be complete by HRE pre- release v0.5.

Q: Don’t dates in Java only go back to 1969?

▪ A: HRE does not use Java date encoding. We use our own method, allowing dates back to 99,999 BC, to an accuracy of one second. This allows the accurate setting of multiple birth order, or the order of Events occurring on the same day.

Q: In the Server mode, can more than one user access the database at the same time?

▪ A: Yes: multiple researchers can be operating on the same database at the same time. There will be some lockout restrictions as results are updated. Otherwise, each record has an ‘ownership’ and last updated timestamp (unlike TMG where only the last updated date is saved for all the records of a Person). HRE keeps a record by record update and the database will keep a record of recent past values of a record so you can trace when a change has occurred and by whom.

Q: Given the Client/Server mode, can I share a common project?

▪ A: Yes. It could be that you share a project at your home and access it from your laptop or your phone when, say, at a library.  Or you may want to share your database with several family collaborators.

Where you place the shared database is the real question. An institution like a museum would probably house it amongst their own servers and share it from there. If you have access to your own shared server, this should be quite easy.  If you don’t, there are a variety of commercial organisations that can provide one.  We hesitate to use the word ‘cloud’ here, since many organisations that offer cheap so-called ‘cloud’ services don’t allow users to run their own database servers.  In order to run HRE in the client/server mode, you would have to choose a commercial supplier that allows you to run H2 server code.

Q: On the Testbed Person screen, it showed person, mother, father, and number of children. Could the wife or wives also be added there? This is very helpful when there are multiple people with the same name.

▪ A: It is intended that these things can be done by settings of the configuration of the particular screen. The internals of HRE are like a set of building blocks which join together and can be configured to do different things through a common component – so almost anything (that make sense) can be joined together. Thus there is no reason HRE couldn’t put a listing of wives in this same list.

Q: On the Testbed Person Relatives screen and in the Zest views shown, there were no spouses visible. Is spouse display not allowed?

▪ A: Remember this was a Testbed. It was built to test the inter-operation of various chosen technologies, not to deliver final-form screen functions, so many parts have not been coded to their full depth.  The Zest technology was used just for the purpose of demonstration, and we have not explored it in depth yet. The final version of HRE may  use something different.

Q: Will there be a way to color code or mark direct ancestors in the picklist and other areas so you can easily see which of say twelve children is your direct?

▪ A: HRE has stepped a long way past where TMG went with Accents. First, the number of color codes is much larger in HRE. Second, all color coding is driven by either a Flag value or a Theme. There are numerous ways you can color-code – it’s really a case of “can you construct a Flag to mark the things you want to color”?

Q: Will HRE be a single file install? What about the database required to run it?

▪ A: Yes – all 5 flavours of the Testbed are single file installs, including the database software. The only exception (not entirely resolved yet) is whether we build-in Java as part of the install package or whether we ask you to install it separately – we may do both. The issue here is to ensure the Java run-time environment is compatible with the HRE package.

Q: Instead of using the generic Census tag, I created Census1800, Census1810, etc. Will these user-created tags carry over?

▪ A: Yes. Although it depends on whether the conversion rules for complex user-defined tags can be systematized. We may have to leave a message for users saying “we don’t think we’ve done this well enough, please check”.

Q: Will the concatenation feature of TMG be carried over?

▪ A: Yes. The principle is that any existing TMG variable or construct will flow over to HRE and will be converted to its equivalent in HRE.

Q: Will it be possible to list all events of a couple, their children and siblings in a single report or timeline?

▪ A: Most probably. In principle there is no reason such a report could not be produced. TMG had a collection of ‘canned’ reports with little control over their content (apart from sentence construction and the report setup options). HRE reports will be written in a scripting language (Jython) using the same substitution techniques as in sentences. So a user could copy an existing report, edit it and create an entirely new report. Such ‘new’ reports could then be exported for use by others.

Q: TMG’s sample database does not seem to include Sentence Variables in Memos. Is there a plan to test that feature?

▪ A: All areas where there is a substitution of any kind will go through a common process of substitution. Whether an error message, a memo with embedded variables, a citation, a name style template, a sentence for an associate or a relationship, event or task – all will pass through the same parsing and substitution code.

Q: In TMG, I wanted the ability to produce narrative that didn’t sound canned. I wanted some variation of, for example, the birth sentence so that every person’s narrative did not sound the same. Will there be some method in HRE, perhaps interchangeable sentence parts, that allow us to vary our sentence structure for each event?

▪ A: In English and other languages it is possible to change the order of phrases and not change the meaning of a sentence. For example the sentence “Person1 married Person2 at a Location on a Date” has 3 components: (1) Person1 verb Person2; (2) preposition Location; and (3) preposition Date.   But the order (3) (2) (1) also generates also a valid sentence – in fact there are up to 6 possible sequences. The HRE Sentence Style allows the user to define a number of sentence templates at the Tag Definition level. At the Event input stage, the user can specify which sentence template is to be used for this person’s sentence. By choosing the template differently throughout the project a user can avoid output text repetition.

Q: Will there be a Relationship Chart feature for two people, such as the report in TMG?

▪ A: Any TMG capability seen as useful will appear in HRE. We see no reason to take away from TMG users any feature.

Q: Has anyone asked Bob Velke what he would have done differently (especially with the data model) after various releases?

▪ A: No. Bob started his design in DOS, then moved to Windows and he was constrained by various legacy components. We had the advantage in HRE design of stepping back and looking at all these issues. Bob would surely have done things differently if he was starting from scratch again, but he did not have available to him the more modern technology which is available to us.

Q: Is the project using GitHub as a repository?

▪ A: Yes. The repository mainly has documentation currently, plus the Testbed code. We are closing the specifications for the database structure shortly and that will then be loaded as well.