What the Future Holds (part 2)
I am adding this discussion to respond to people who posted on the "What the Future Holds" because that discussion had been closed."
Up Front: I don't buy the argument that the old technology used by the original Source Linker will cause it to fall apart. In my career, 40+ years career in the software industry, I have never heard of this. The most usual and valid reason is cost. In most cases the cost of simultaneously supporting an older version is very small.
Áine Ní Donnghaile: Thank you for your help and replies. My position is not about when the change take effect, it is about how the changes are deployed.
Here I would rather use WWW and software development DeFacto standards. The change announcement should use a rapidly and accurately understood icon. Unfortunately, there does not seem to be a de facto standard. This does not absolve me of being guilty of poor comprehension of visual clues. I should have seen the information.
The industry standard for releasing software is Beta, and some variants. The beta test should include some recognized experts, and as many users as possible. Another approach is to provide support for changing levels into the software package itself. When suggested, a software supplier usually cites increased investment and support costs. Valid reasons for most suppliers. But not for FamilySearch.com. It has already made the investment to provide users the best flip/flop solution I have seen. Support costs can be avoided by making the older software "as is". No support is not free, but it can be reduced to almost nothing. The next best solution is to document solutions for users.
PamelaCarter: Sorry I did not make it clearer. I am not unhappy with the new source linker. Just the release style.
The "Next Gen" Source Linker will become an outstanding application, but it will require time and more releases. I want to take the pressure off the developers; they will not have to spend their valuable time dealing with a no-further-choice release.
I am glad the ROC and VROC Missionaries have contributed to the rollout of the new Source Linker. Experience makes me estimate the amount of testing the new Source Linker should undergo much more.
Gordon Collett: First paragraph was addresses in my first response above. I don't buy it. However, FamilySearch.com does. They make the call. They will go ahead.
The rest of your post is a prediction. I sincerely hope you are 200% right.
We'll see soon.
Douglas McPhaden
コメント
-
Source Linker is some of the oldest programming on FS's website. It was introduced a full decade ago (https://www.familysearch.org/en/blog/a-new-way-to-attach-sources-to-family-tree-the-new-source-feature-part-2-of-2), so I can well believe that it needs to be re-done. For example, in 2014, source tagging didn't exist yet, so it had to be tacked on later, which is never as efficient as building in a feature from the get-go.
The other thing to note is that you're coming in at the very tail end of a years-long process. They did first put the new version on the beta site. (The problem is that most of the indexed records database is not duplicated in beta, so the testing possibilities there were limited.) This group has existed for seven months now, and as I recall, it wasn't created immediately when the option to try the new version first showed up in production, so we've had well over half a year to try it, test it, report the bugs, ask for features, and so on.
I really don't understand what your objection is to the "release style".
2 -
I go along with your point:
I don't buy the argument that the old technology used by the original Source Linker will cause it to fall apart.
FamilySearch engineers usually present arguments about obsolete fonts, etc. causing certain changes to be essential, but some of the websites I use have seen no noticeable change in over ten years, yet they still manage to keep functioning.
I'm perfectly happy with enhancements that will improve the quality / functionality of a feature / page, but I just can't adjust to some of the changes made in the last couple of years (specifically, the revised Records page - hopeless for presenting / examining census records) and I'm finding it difficult to adjust to the new source linker page, too - which is why I've put off using it until just a couple of weeks before the old one is due to disappear!
To be clear - I'm all for anything that provides improved functionality, but just don't like the fonts / typefaces that the FamilySearch engineers are choosing to use. Can't illustrate with screenshots at the moment, of course, but why do these have to be so different to the ones being replaced? Everything is so BOLD and "in my face" - the only comparison I can think of is, I find it's like reading a page from a magazine instead of one from a novel!
0 -
For what it's worth, the statement in the Community News says:
"… the Source Linker has been undergoing some updating because the older Source Linker is running on technology that has reached its end of life and has to be replaced. …"
I read that as referring to the hardware - which made little sense to me as I've never seen application software that couldn't be replatformed. If, on the other hand, it refers to software infrastructure, that's a different kettle of fish.
So the statement lacks clarity.
1 -
For some background on the need for changes, see Todd Powell's 2023 RootsTech presentation: https://www.familysearch.org/en/rootstech/session/whats-new-at-familysearch-for-members-of-the-church-of-jesus-christ-of-latter-day-saints-2023?lang=eng starting at 21:15.
For older background about previous changes in the website, this is an interesting article: https://www.datastax.com/enterprise-success/familysearch and the video is very interesting in regards to one of the platforms FamilySearch built on. Couldn't find a date for this article.
Regarding end of life technology, I've always assumed this was software. A problem with FamilySearch through the years that gets mentioned peripherally here and there is that they always seem to be pushing software platforms, like the Cassandra platform mentioned in the article, two or three steps beyond what it was ever designed to do. Maybe, if they were even still using Cassandra, FamilySearch ran into these types of issues: https://www.scylladb.com/2022/02/14/why-14-teams-moved-on-from-apache-cassandra/
As a last bit of musing, I wonder if this React: https://en.wikipedia.org/wiki/React_(JavaScript_library) is the React that Todd Powell mentioned that will let FamilySearch be much more nimble with developing and correcting problems in the website.
3