4 ms·

Brian from Backblaze here. > Backblaze has repeatedly refused to add something as simple as pagination to their B2 web interface Not "refused to add", it's on

Brian from Backblaze here. > Backblaze has repeatedly refused to add something as simple as pagination to their B2 web interface Not "refused to add", it's on the list! We have a relatively small team of programmers (no VC funding, no deep pockets, we can only hire when we can afford it), and every week we meet and set priorities and the engineers work on the very highest priority stuff in order. The pagination has been delayed in favor of other features and bug fixes. I'm not saying we prioritized correctly, I'm just explaining why it has not been done yet. > which most people would not dream of going live without Part of the challenge (fun?) of building a business is the "build order". We don't make any money until we release the product, so as soon as we can justify it we released it in "paid beta" to collect feedback like yours and make some money which in turn allows us to hire one more programmer and build that feature for you. If B2 doesn't fit your needs yet, my only request is that you come back every six months and check again, because it is always evolving.

cookiecaper 9y ago
Thanks for the reply. I didn't know that Backblaze was bootstrapped and I really do respect that. But this has been "on the list" for about a year. For an object storage offering, crashing the tab whenever someone opens a large bucket is not something to brush off. I was initially sympathetic, but since B2 also does not have an API call that will allow me to update my filenames (either 'rename' or 'copy to new name') to comply with the "virtual folders" feature, which either didn't exist when I wrote my script or which I overlooked, I'm stuck. I asked support to make a bigger push to get some simple pagination added and they said they'll register the feature request. I emphasized that this isn't just a convenience feature, but that it actually makes the tab too slow to use and/or crashes it (don't remember which right now to be honest), but it didn't matter. I asked support to make a one-time exception and find someone who could run a find and replace on the object filenames to comply with the new name structure and they declined to do so. The only option is to pay to download the incorrectly-named files and reupload under new names, which would cost a lot of money in download fees, not to mention a lot of clock time and bandwidth. Of course, this isn't really intractable. To avoid extra charges from Backblaze, I could just upload the files again since I still have a local copy (for now), but that would also take a lot of time and bandwidth. I could write a client-side script that would stop the B2 interface's broken behavior, and that would make the bucket semi-browseable/usable via the web. But I just don't feel a need to do all that when I could use other solutions that are better about this kind of thing. B2 left beta a while ago iirc.

cookiecaper 9y ago

Thanks for the reply. I didn't know that Backblaze was bootstrapped and I really do respect that. But this has been "on the list" for about a year. For an object storage offering, crashing the tab whenever someone opens a large bucket is not something to brush off. I was initially sympathetic, but since B2 also does not have an API call that will allow me to update my filenames (either 'rename' or 'copy to new name') to comply with the "virtual folders" feature, which either didn't exist when I wrote my script or which I overlooked, I'm stuck. I asked support to make a bigger push to get some simple pagination added and they said they'll register the feature request. I emphasized that this isn't just a convenience feature, but that it actually makes the tab too slow to use and/or crashes it (don't remember which right now to be honest), but it didn't matter. I asked support to make a one-time exception and find someone who could run a find and replace on the object filenames to comply with the new name structure and they declined to do so. The only option is to pay to download the incorrectly-named files and reupload under new names, which would cost a lot of money in download fees, not to mention a lot of clock time and bandwidth. Of course, this isn't really intractable. To avoid extra charges from Backblaze, I could just upload the files again since I still have a local copy (for now), but that would also take a lot of time and bandwidth. I could write a client-side script that would stop the B2 interface's broken behavior, and that would make the bucket semi-browseable/usable via the web. But I just don't feel a need to do all that when I could use other solutions that are better about this kind of thing. B2 left beta a while ago iirc.