2023-07-20

Fool's errand in pushing the envelope of software limit

$

Fool's errand in pushing the envelope of software limit

I am not unaware that a lot of my time was spent in at least the last two weeks testing the limit of software, specifically Obsidian and later Logseq, to see how many files such local-file-based systems can handle and still run relatively smoothly.

I am greedy in that I also want the convenience and power of sync, so I put files on Apple's iCloud, which is honestly notoriously slow and finicky in performance. Unfortunately it is so popular, more so than, say, Dropbox and Box, that it's the only free cloud solution that these two PKM tools support.

So, the question of answer each of these tools' file limit is further complicated by the sluggish and unstable nature of iCloud, so that the whole experiment is cloudy and becomes an entangled mess.

Today, after more such frustrating experiments trying to find the upper limits of files in deciding which tool is best for housing my deal glossary/terminology management and lookup system, I have come to this realization: Stop chasing after the unreachable and hoping the software in question will do what I want instead of what it is technically capable. Software design is bound by many factors and these are competing with each other. For example, if Obsidian and Logseq boasts of local files then they are more likely more constrained in the number of files they support without crashing or hanging, as I witnessed clearly. On the other hand, Anki is a perfect example of a database solution, and therefore excels in supporting my current 1.2 million entry database of myriad glossaries without batting an eye! What would be ideal is the capacity of Anki and the modern features of Obsidian/Logseq/Notion/Tana! But such a perfect tool doesn't exist and no amount of money can buy it realistically. (Of course, I can fund the development of my ideal software to serve my demand, but who am I kidding? I will make a terrible product or project manager!)

Bottom line, just stop. Perhaps the best medium-term approach is to be happy with the less ambitious Anki data merge project, and be satisfied and content in enjoying the not-trivial reduction of records from 1.2 million to under 800K! If I can achieve this goal, and enjoy better speed on Anki, I will have already won. In the process, I have also gained valuable knowledge about how to design and structure glossary data, and be ready to do something big when software products emerge that meet my need of capacity and performance. (Think Tana, but I will not hold any delusions.)

Okay, enough ranting. Now dial back the monkeying around and settle for something realistic that will result in actual tangible improvement of my professional and personal life.