Newest pre5 unusable for years under Linux(s)

Subscribe to Newest pre5 unusable for years under Linux(s) 2 post(s), 2 voice(s)

Avatar mikebutash 3 post(s)

So I’ve used jbidwatcher for a very long time now (8-10+yr?), and the last 2.99-pre5 with ruby scripting enabled broke it from working on like every linux I’ve used, now for years with different distributions, including Ubuntu and Arch Linux.

I’ve loaded jruby and about every other lib I know to try and fix compatibility, but under linux your jar doesn’t seem to look in the right place for whatever it wants. The debugs are vague, but in any regard, jbidwatcher doesn’t launch and doesn’t work in this version under linux since enabling default the scripting. I use an older version (JBidwatcher 2.5.6-11-gac197d6), highly dysfunctional, but it’s otherwise usable for snipes as long as I don’t expect it not to lose the result after successful bid.

I can’t use 2.99-pre5 due to ruby dependencies, so using the prior version under derby db randomly loses/adds/corrupts entries every time it gets restarted. Not sure if this is Derby broken or JBidwatcher. Pre-auction works, keeps cost, snipe, and will snipe successfully. If I add an auction and do not bid, it will randomly move to complete, or not, which is more often the case. I think meaning it’s not updating the DB, the auction is more often just gone after completion.

Even more odd, occasionally I restart the desktop or system, and it auto-starts, with like 50 duplicate historical entries of the same things, sometimes very old from my historical searches. Sometimes it starts with nothing but one or two random old entries. It’s rather perplexing what would cause such random behaviour for like the past 2 years across both ubuntu AND arch linux using mostly the same java 1.8’s.

It seems I’m the last person on earth using linux with jbidwatcher (only reason I use anything java, period – it should die, please.), so for what it’s worth, jbidwatcher hasn’t worked properly in some time now. It would be nice to see it work again, or a successful port to something not oracle-owned and molested (and forgotten).