I run a coupla TTS apps in Puppy. They're all Windows apps, running under WINE.......TTS Reader, Balabolka, and what was at one time the "industry standard", TextAloud!
You have to install a few Windows items into WINE, then you need to install voices as part of the "back-end" stuff before installing the apps themselves. The good news is that once everything's in place, these do work very well.......even though these are the older-type, less 'natural-sounding' voices than what we have available today.
I found a free source for some of the AT&T "NaturalVoices" from a few years ago. Do be aware that these are NOT small packages, and the average "voice" can run out anywhere between 600-800 MB, each
Although you can find equivalents for pretty much anything on our side of the fence, text-to-speech is one of the very few areas where Linux is NOT well represented, and Linux TTS software is nearly non-existent. A few open-source 'engines' exist, though they need to be 'trained', and to make a usable package you still need to add a lot of items yourself.
is right, I'm afraid; Windows has this particular facet sewn-up, and all devs who cover this area do so exclusively for Windows. But most Windows TTS software runs well under WINE.....with one exception; don't bother with Dragon NaturallySpeaking, because they've made good and sure that their paid-for software will ONLY work properly under Windows itself. And an annual subscription costs an arm & a leg.
(Unlike some Linux purists, I have no issues with using WINE. I will run whatever mixture of software it takes to achieve the end result I'm after. There's no shame in using Windows software; no denying that for some things, you simply can't beat it. For other stuff, Linux will run rings round it. At the end of the day, it's a simple case of using whatever works for YOU.)