Colleague takes a holiday and I make him happy by playing his replacement. He sent me a ??250,000 TUs TMX file.
I've opened it in CafeTran, removed all duplicates and segments that only contain numbers: 3 secs. Up and running.
Now I have to import it in Studio 2015, in order to make a Trados Style analysis.
Oh man, this tool is so incredible slow.
Hmm. Yes, and no.
It might be slow, but that's because it is actually indexing the TMX, rather than merely opening and holding it in RAM. ‘Indexing’ might not be the right word, but the point is that it is adding it to a special database of some kind. However, once you've imported it the first time, the next time you open a Studio project with the same TM attached, it will be instantaneous. It will most likely also work faster as you work than in a CAT tool that doesn't index its TMXs and works on them directly. For example, in Studio, I am able to translate with much larger TMs attached to my project than in CT. I also think that Studio is a bit snappier than memoQ in this regard.
Shall I take a power nap?
Not to mention the dozens of error message that I get all the time.
Regarding verifying: you don't actually run QA, do you? ;-)
On a serious note: I'm getting worried by the many error messages that I get when processing clients' packages in Studio 2014 or 2015. Perhaps these are caused by my upgrading my Windows 8.1 VM to Windows 10? That'd really be a pity.
Incompatibilities like these caused me to start using virtualisation some years ago. (Trados Translator's Workbench, requiring a specific Java version.)
Seems like a big deal, but Studio is actually already very fast. It would be cool if there were also a dedicated 64-bit version though.
When working with it, with very large TMs connected, and tons of separate TBs (some very large), working in the editor is snappy, and moving to the next segment also very fast. No waiting for stuff to load. Sure, some stuff is a bit slow, but as I already mentioned, there are reasons for (some of) that. TMX importing, e.g., is slow because it is importing it into a special database, which will allow you to work very fast later.
Yes, Michael, I do understand your way of working. But mine is different. I use one TM per project. And in the cases where a client sends a legacy TMX file, I want to start working with it right away. Do not want to do any import stuff. Ain't using Total Recall either.
It takes me all of 5 minutes to prepare a legacy TM from a client. And, once it's prepared, its immediately available for future use. No matter how big.