At long last, Skype returned back to normal operation today, and has been stable.
Why on earth it took so long is a question that only the developers could answer, or why on earth the issue came up to start with after a long period of reliability, and no significant changes occurring that are obvious to the user, and even then, if a significant change occurred, they could have just reversed it.
Why would such a problem take so long to fix or rollback from ?
They blamed algorithm deficency, but again, how does that stop them from reversing to the ‘earlier algorithm that worked’.
Was Skype’s software time bombed in some way? Or did they find another limit between 16 and 32 which is to blame (assuming they use epoch in that algorithm).
I guess the key question from me is, what on earth is an algorithm deficency that took so long to fix, and couldn’t simply be rolled back from?
It lost eBay more than $1 billion dollars – that’s not a bad thing :).
What ever happened to coding revision storage, so that at least a backup of some latest version of the code is somewhere?
What about the developer who sold Skype, did they bother to get a hold of him/her?
It doesn’t really add up to me, unless of course the complete algorithm is dynamic, and it had an upper bound limit that was unprotected, and they couldn’t find any way of countering the upper limit, and this is used during sign on, and the upper limit was based on something dynamic – such as time?.
The reason I resort to time here is because such an algorithm would be needed to be based on something time like to have the effects that it did, or have something to do with time, such as system loading data over time, but then, a reboot would bring most of the network back online.
Attacks would be more likely, such as a character in a username field, or something like that, but they ruled those out themselves.
Anyway, it’s fixed now, so it’s only made Skype’s software stronger as a result of it (unless they are dirty script kiddies and just put a little hack over it).
In other news, If you ever looked at Telstra as being there for Regional and Rural consumers, think again.
Telstra have been caught out again, with Long Line ADSL.
It was originally proposed to be enabled in 2005 at nearly 250 locations, by a former Telstra head, Doug Campbell (no questions why he is no longer at Telstra).
Long Line ADSL brings line lengths closer to the home, and extends the capability by 20km.
Telstra badly wasted more shareholders money on 200 devices sitting in warehouses, not getting deployed, because Sol would rather push NextG, than affordable, innovative options like Long Line ADSL.
Long Line ADSL is available in 50 areas. It’s not something Telstra want you knowing about, removing every reference to it from all over its website to ensure that few people know about it.
However, despite the attempted coverup, there’s a thread on it over at Whirlpool, and ITWire certainly kept a pulse on it also!
The sad news is Telstra would waste money on buying 200 devices capable of delivering broadband to many users, as well, would waste the money buying ADSL2+ technology, and supplying the several times superseded ADSL1 technology, and even then, limiting its capability.
Telstra aren’t considering any consumer here. The sooner the generations signing with them realise this and go elsewhere purely to show them that consumers shouldn’t be held to ransom, the better!