not sure if this fits into the Expansion pack or the Bug/Testing Forum.
Some here said they would repair/rewrite/adopt old and/or broken oxps.
After long experimentation and putting oxps back in one by one I have found that under Windows Vista and Oolite 1.70 the following oxps seem to have one or the other serious issue:
Each of them needs the "homebrew parser", supposedly discontinued beyond 1.70, or gets my system to somehow loop and use up ridiculous amounts of RAM (I have 2GB RAM and Dual-Core, Graphics Card NVIDIA GeForce 8600 M GT) until it crashes.
Perhaps this helps with the choice? And I would be very happy if the issues could be resolved - trying my hand at that myself is impossible, as my last programming experience was with Basic & Pascal in the 1980s ...
Cheers
Lestradae
Last edited by Lestradae on Fri Feb 22, 2008 9:20 am, edited 2 times in total.
Is it safe to assume that, after removing these offending OXPs, Oolite runs without any further problems on your system? I sure hope for a positive answer.
Is it safe to assume that, after removing these offending OXPs, Oolite runs without any further problems on your system? I sure hope for a positive answer.
Somehow I doubt it If he has all the oxps. In combination with the memory leaks of 1.70 this is pure suicide.
Somehow I doubt it.He has all the oxps. In combination with the memory leaks of 1.70 this is pure suicide.
I know ...
I think it's safe to say that Oolite is able to read them happily using the new parser so future versions will be able to read them.
A future parser? Didn`t Ahruman originally say that he would discontinue the homebrew parser?
Is it safe to assume that, after removing these offending OXPs, Oolite runs without any further problems on your system? I sure hope for a positive answer.
It now runs much, much better. Still testing. Will report here success or failure as soon as I`m sure about that.
when he sees that list kaks might start to regret saying he had a lot of time free !
Perhaps he should be awarded the commie title "hero of work" if he turns to that list?
Noted that now, after taking out the aforementioned oxps, suddenly Trident Down 0.5 starts making trouble.
Just as a point of clarification, I don't expect 1.70 to run flawlessly all the time, given the existence of memory leaks, various bugs etc. The point of my question was to assert that Lestradae does not suffer any more from the crash havok that was taking place on his computer before the removal of the problematic OXPs. Which, thankfully, seems to be the case.
Just as a point of clarification, I don't expect 1.70 to run flawlessly all the time, given the existence of memory leaks, various bugs etc.
Me neither.
The point of my question was to assert that Lestradae does not suffer any more from the crash havok that was taking place on his computer before the removal of the problematic OXPs. Which, thankfully, seems to be the case.
Just as a point of clarification, I don't expect 1.70 to run flawlessly all the time, given the existence of memory leaks, various bugs etc. The point of my question was to assert that Lestradae does not suffer any more from the crash havok that was taking place on his computer before the removal of the problematic OXPs. Which, thankfully, seems to be the case.
Both Falcon & swift are 90% sorted: they don't generate errors at startup, but still cause random crashes from time to time. Fingers crossed, I might be able to figure out what's happening in the next couple of weeks!
Orb: Charlie's going to add a thing or three, then we should be able to release a much better & bigger oxp, fully compatible with 1.70
Will probably have a look at all-stars whenif I get the spare time!
The point of my question was to assert that Lestradae does not suffer any more from the crash havok that was taking place on his computer before the removal of the problematic OXPs.
For the first time in months everything running completely smooth, no jerking, no crashes, and the stderr doesn`t show a single error! And that with (still!) 147 oxps in!
Oh Dear. I guess I asked for it!
Long live Kaks, our hero of work! *waves red flag*
Both Falcon & swift are 90% sorted
Are those 90%-versions already on the wiki or should I better wait for the moment?
Orb: Charlie's going to add a thing or three, then we should be able to release a much better & bigger oxp, fully compatible with 1.70 Will probably have a look at all-stars when if I get the spare time!
Very cool! And, seriously, thanks for all the work you invest into this great game!
I think it's safe to say that Oolite is able to read them happily using the new parser so future versions will be able to read them.
No. The warning means that a plist can’t be read by the “new” parser (which isn’t actually new)…
To clarify (again): the functionality of the “homebrew” parser has not changed. It has never worked very well; it doesn’t work with all valid property lists, and while it “works” with some broken property lists, most of those property lists do not make sense and what the homebrew parser reads is mostly nonsense. Those plists have never worked as intended with any version of Oolite. Mostly this has to do with plists that have two <key> entries in a row, or skip a <key> entry, at which point the homebrew parser starts treating <key>s as <string>s and vice versa. The homebrew parser has been updated to notice this and produce warnings, but it still passes the same confused mess up to Oolite.
The message says the homebrew parser will be removed “in a future version of Oolite”. My plan is to keep the homebrew parser (and the warning) until the MNSR, then remove it for an accompanying “developer” release (which will have the same debugging features as current test releases). Hypothetical test and stable releases after the MNSR will not include the homebrew parser, because perpetual backwards-compatibility with OXPs that have never worked properly is not a reasonable goal.
I have found that under Windows Vista and Oolite 1.70 the following oxps seem to have one or the other serious issue:
I know ...
Sorry for the misunderstanding 
