Hi Joris, Thanks again for all the work you've been doing for the community!
For the mailing list archive, the obvious and easy solution to migrate away from your archive would be to just rely on the mailman archiving that is used for other OSGeo-hosted mailing lists (like https://lists.osgeo.org/pipermail/gdal-dev/), and that had been disabled explicitly for the tiff mailing list. The downside is mailman is that it isn't really searchable (what Nabble used to work around, but is no longer there unfortunately). Anyway if we had access to the raw emails of the mailing list (do you have that?), I presume we could re-ingest that into mailman, and reconstruct a full libtiff archive. In the worse case, we might just re-enable mailman archiving onwards, and rely on the wayback machine to access older threads when needed (my personal experience is that it is very rare that I need to dig into ancient emails not in my mailbox, but newcomers might perhaps find it useful to browse into past discussions occasionally)
The tag reference and file format are very useful (I've even seen them referenced from official documents). Not sure what the community could come up with. Still in the worse case, we might crowl the wayback machine if awaresystems.be/imaging/tiff would become available. I guess it wouldn't hurt though if you could dump your source code into some github repo though, despite the potential difficulty of making sense of it and the uncertainties on the licensing of some components. At least that could be a starting point.
Even Le 24/04/2024 à 22:17, Joris Van Damme (AWare Systems) a écrit :
Hello, I would like to discuss https://www.awaresystems.be/imaging/tiff.html and everything on there. At this point, I feel the humble and inefficient way I manage and publish some data is not sustainable. On the one hand, I'm unable to properly maintain things as I'm not even home on a regular basis, and on the other hand for proper function of things I do think you all need at least some proper way to access a record of this data, and when it comes to the mailing list, preferably a real-time one. Some people on this list have expressed some interest in the data, in the recent past. However, discussing it further, there always seem to be a few major roadblocks. - all of it is in a proprietary format, defined only by my own code - a mailing list archive of 1870, so to speak, is cute and all, but is it useful? also, I should mention not all of it is very properly formatted (some periods are, some aren't) - much of the tag data, is just... inspired deeply... by specifications, as in my mind it should be, because mostly that is the correct information to put out there. but there is this small issue of the real world of copy... whatever. I really don't care much, but few people approach it that way. Are there any thoughts on this? I just updated the mailing list archive. In an ideal world, instead of update things in the same clumsy way again in six weeks time when I'm home again, I'd like to just close down the above mentioned at that point. If there's some perceived need to preserve some of the data in some way, I can put in some final effort to export in some way if it's not too much of a hurdle. Joris
-- http://www.spatialys.com My software is free, but my time generally not. _______________________________________________ Tiff mailing list Tiff@lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/tiff
