[X] My Assistant

With the new resampling system, the H@H Downloader will currently only download original images for new galleries, rather than the old resamples. While galleries uploaded prior to a couple of days ago will still be using the resamples from the old system, those will be phased out over the next few months in favor of resamples from the new one.

Similar to the proxy functionality, with the addition of static ranges, using the Downloader will likely be detrimental to a client if the cache is close to full. And with the resample change, this effect will be exacerbated since the files downloaded will tend to be larger.

As such, if the H@H Downloader functionality were to remain, I would have to rework it to use the new resampler. While this would probably be feasible, my preferred solution would be to phase it out entirely and instead make it so that running H@H gives you free archive downloads similar to what you'd get for donations.

The archivers have plenty of capacity, and this both reduces complexity and completely avoids the problem where files downloaded with H@H will overflow the cache and make it discard actually useful files. I am also considering adding an archiving option for at least one tier of resamples, probably the 1280x one, for people who don't care about "absurd resolution" images.

So, for those of you who use this, would there be any reasons why you would want H@H to handle the downloading rather than have it give you free archive downloads instead?

my preferred solution would be to phase it out entirely and instead make it so that running H@H gives you free archive downloads similar to what you'd get for donations.

I'm fine with this.

I am also considering adding an archiving option for at least one tier of resamples, probably the 1280x one, for people who don't care about "absurd resolution" images.

This is a waste of time, don't do it. What is the point of archiving a non-optimal fileset anyway? If someone is really so adamant about collecting garbage let them do it themselves.

This is a waste of time, don't do it. What is the point of archiving a non-optimal fileset anyway? If someone is really so adamant about collecting garbage let them do it themselves.

People keep requesting it, and it would restore a way to easily get resampled files, which H@H lost for new galleries with the resample update.

People keep requesting it, and it would restore a way to easily get resampled files, which H@H lost for new galleries with the resample update.

Seems crazy to me, but I guess someone out there wants it. Also if you do this will there be a separate pool of free archive downloads for these resamples or will it share the normal pool?

If you're going to remove the downloader it just occured to me that currently GalleryInfo.txt files only exist when downloading through the H@H downloader. Any chance that the archives could get this file? I use it to sort my downloads into categories.

Why not add archives download to the API so that people that used to automate the downloads using H@H can still do it?

I use the H@H Downloader in order to fill my H@H caches with popular images, but it does not really work anymore since the static ranges anyway, so I won't mind its removal.

Why not add archives download to the API so that people that used to automate the downloads using H@H can still do it?

I didn't even know there was an API, archive downloads should be added to the API for sure.

It would depend on the hitrate of the clients, as it does today. I would try to set it so that, considering it adds the option of getting originals, you couldn't justly make the case that they lost anything.

As part of that whole thing I'd probably change the free quota from a number of archives to a number of MB, converted based on what the average downloaded archive size is today.

As long as the download quota does add up (up to an upper limit, i suppose to be realistic).
My typical use case for hathdl was: Get a good amount (>50) of hathdl files from galleries, upload them to H@H, wait a day, compress the folders and grab them via SCP. Repeat 3-7 days later.

This post has been edited by simrock87: Mar 28 2016, 20:12

that would make it easier to get hath if we had an easy to set up and use H@H client like that of vuze or bitcomet. i still don't understand how to set it up. I wish there was just a open installer and sign in to get hath from h@h. It would normalize hath prices too.

that would make it easier to get hath if we had an easy to set up and use H@H client like that of vuze or bitcomet. i still don't understand how to set it up. I wish there was just a open installer and sign in to get hath from h@h. It would normalize hath prices too.

It's not difficult, it's literally 5 minutes of work if you're using linux and screen to run H@H in a separate session (and I pretty much suck at administering linux), otherwise even less.
If someone has trouble configuring the network or the OS chances are that his client will not be a great contribution and possibly "steal" requests from other, higher quality, clients, thus possibly hurting the network.

It's also not possible to have is as easy as vuze or bitcomet (which, by the way, aren't even the best torrent clients available, bitcomet being the worse offender of the two) because no torrent clients require a public IPv4 IP or an open port (they can start the connection, use NAT traversal etc etc).
Very different things.
You should really run H@H if you know what you are doing (every failed request is hurting one user's experience).

This is a waste of time, don't do it. What is the point of archiving a non-optimal fileset anyway? If someone is really so adamant about collecting garbage let them do it themselves.

Just because you think it's a waste of time doesn't mean everyone else thinks the same way. There's more than enough galleries out there with absurd resolutions and filesizes that people without equally absurd monitors can't properly display anyway. I hardly know anyone who really needs images in 6000 * 9000 resolutions like this guy often uploads:
http://ehentaihip.com/uploader/shakuganaexa

What I like about the current H@H downloader in particular is that it allows me to easily do browsing from one computer and downloading from another. I run H@H on my desktop, but I tend to browse the galleries on my laptop. My laptop browser is configured to have my hathdl folder on my desktop set as one of the default download locations, so I can just grab the .hathdl file, save and continue browsing. It'd be a fair bit more tedious if I had to download the galleries through the archiver and copy them up. It wouldn't a major annoyance, but I certainly find the H@H downloader especially convenient for my use case.


If it were possible to do something similar with archive downloads though, that'd be pretty boss.

Seems to me that you'd just need a download manager with a watch folder or some kind of context menu to submit the download link.

No, no, no. Please don't remove this. It's pretty much the only reason I use the site atm! As someone earlier mentioned, it's just so easy to browse galleries, save the hathdl files and leave your computer to churn away doing the downloading. If I have to manually download stuff I'll probably give up H@H altogether. I download about 50-100 galleries a day using this system. Manually downloading that many galleries EVERY SINGLE DAY makes my head hurt even thinking about it.

Using the H@H downloader is a much simpler and less intrusive process. It'd be a shame to see it go.

Manually downloading that many galleries EVERY SINGLE DAY makes my head hurt even thinking about it

It does indeed.

Manually downloading that many galleries EVERY SINGLE DAY makes my head hurt even thinking about it.

It does indeed.

Adding the download functionality to the API would solve that, though.

There have been a couple of times that the H@H client was the only way I could download particular galleries. I have no idea why; the BT wasn't available, and the archive downloads didn't work. I think retaining the feature if it didn't require significant rework would be preferable; otherwise, well, I guess I'm fine with it providing free archive downloads.

i am in favor

I have some issues with it.

• I like having that galleryinfo.txt it helps me a lot.
• I like having partial gallery downloads. Example: Sometimes I only download 30 images in a 2000 image gallery.
• The hathdl files allow me to have a low size backup that allows me to queue a restore of all the galleries I had downloaded in case where I have my galleries crashes (it already helped me twice).
• hathdl files allows me to queue downloads. I don't really care it takes a night to download a 200 page gallery, tbh.
• Having larger files are not an issue to me.
• Adding extra metadata (e.g. custom tags) to the hathdl file that, that are later added to the galleryinfo.txt, help when searching for galleries using windows search. That is really useful to me.
  Sometimes, the gallery description contains useful data. The most common being links to different chapters and on which page each story starts (based on the author's name)

Sorry for taking so long to answer

This post has been edited by Jay Low: Jul 8 2016, 02:10

Used to use the downloader to populate the cache.
If i download a gallery (which is rare) it is either via torrent or the Archiver.
So in favour of removal (indifferent).

This post has been edited by Knabber96: Jul 8 2016, 21:20

Thanks for the feedback all. The Downloader has now been replaced with a resample option on the archiver as well as a free quota for people who run H@H. You can read about this here.