The recent news that the patent dispute over LTO-8 has been resolved is a welcome one. But the time it has taken to sort out might mean that it is too little, too late.
As we discussed as recently as June, there's a problem with the supply of LTO-8 tape. The ability to store 30TB of data to a reasonably reliable archival medium at 360 megabytes per second is a rare, important and useful thing, so this shortage was not well received, especially as it wasn't ever a technical issue – it was a legal one. LTO-8 tapes worked fine, could be manufactured, and had been distributed, but the world was denied this little bit of progress because of patent law.
The story today is therefore not really that Sony and Fujifilm have somehow pulled their respective irons out of the fire. It'll be instinctive to most technical people to form some sort of opinion on patent fights: those involving a genuine question to answer over a matter of technical interpretation, and this which are more or less solely about people trying to make money. It's very difficult to make that differentiation with the Sony/Fujifilm situation, because very little information has been released about exactly what the disagreement was. The substantive parts of the legal document can be quoted in full:
It is ordered that:
OK, that's great – fully three years after LTO-8 was introduced, can we possibly get back to regular, easy availability of tapes?
Er – no. One thing to be aware of, despite headlines to the contrary, is that this information does not suggest an immediate solution to of the tape shortage. It'll take quite some time for the production and distribution apparatus to spin up, and then more time yet for overdue backorders to be fulfilled. The thinking seems to be that it'll be early 2020 before LTO-8 tape becomes easy to get. That would be one thing, but by rights, the next generation of LTO tape is due next year anyway. There has been a two or three year gap between generations going all the way back to LTO-1 in 2000, and LTO-9 would, by that measure, have been due before 2021.
LTO-9 should be highly capable. It fulfils the original goals of LTO, which were to ensure that each successive format would have twice the capacity and twice the speed of the previous one. That hasn't always quite been achieved, with speeds lagging slightly; LTO-5 was almost twice as big as LTO-4, but only about 1.14 times the speed, and there was a similar story between LTO-7 at 300MB/second and (on paper) LTO-8 at 360. LTO-9, however, is projected to achieve a mind-melting 708 megabytes per second on tape, without helical scan. It's not quite clear whether users of this technology will be required to wear a fighter pilot's G-suit, but to say that it is hugely capable and a massive engineering achievement somehow seems inadequate. LTO is excellent.
Nonetheless, that grinding sound you can hear is the teeth of LTO-8 early adopters mourning a large part of their investment. They've been able to use LTO-7 tapes reformatted to an interim 22.5TB format, but it's slower than 8, and not really what we were promised. The big problem, though, will be if this whole sorry escapade has a negative impact on the takeup of LTO as a whole. It needs to look reliable because it's not cheap – at least, it's not consumer cheap, though it can start to seem very affordable compared to the costs of, say, a professional data forensics lab recovering lost files from a broken RAID. LTO needs to be used in more places than it currently is, not less.
The consortium that runs LTO appear to be well-meaning people. It is a collaboration of competitors, but it is also a hugely positive example of what can be achieved with a bit of lateral thinking. The tape shortage is exactly the sort of situation that the consortium was designed to avoid and with any luck, the people involved will be as horrified by what's happened as anyone else. If all goes well, that will allow them to avoid it happening again. Here's hoping.