Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[greek]

1) smart LowID check at start up

I am wondering, are there two ways to do LowID checks? Why not just pick the best one all the time? Why would the user care if there is a smart and a not so smart LowID check? I suggest sticking only in one of the possible algorithms of LowID and eliminating this option. Programs should be automatic.

2) I.C.H active

Is there any reason why one would not like to use Intelligent corruption handling? Why not have this enabled at all times and remove it as an options?

3) Check disk space

There is no reason to check disk space! If there is not enough disk space, the "status" will turn from "Downloading" to "No disk space" and all downloads will automatically pause.

4) Show percentage/Show progressbar

If I uncheck both of them, the "Progress" column in the downloads tab is empty! I think one of the two should be always visible and I suggest that both of them are visible and those options are removed.

5) Check for new version at startup

Oh come on! This one is just plain teasing!

I think that the gui (and probably the core) should be simplified and cleaned up. Removing some options and automating the behavior is a good start.
Please let me know your opinion.

[GonoszTopi]

Let the user have options. All of them have reasonable defaults, and not touching them makes no harm.

[greek]

Well sure, but if some options can be determined self consistently and optimally by them selves why not do so? Aren't the programs supposed to be automatic?

[lfroen]

I'm not sure what "Smart low ID check" is an option, and regarding others:
* Disk space is an option, because some filesystems support sparse files, and some people prefer amule not to fill all available space (linux don't like "/" being at 100%).
* I.C.H. takes cpu time.
* Show ".." is option because some people want to see %% written on progress bar. If progress bar itself disappear, IMHO it's a but
* Check for new version is doing what it's supposed to do. Don't like it - turn it off, that's why it's an OPTION.

[greek]

Ok I see. So disk space check is useful in linux. My bad. Is I.C.H. really slow? Of course it will take more cpu if it is "intelligent" but if it is not major why not just keep it on all the time? Are there benchmarks that demonstrate the difference in performance? I am just wondering. The progress bar is an aesthetic issue mostly, but on the other hand the aMule interface is not exactly refined, why would anyone bother to offer options about progress bars, percentages and styles, when, for instance the skin support is incomplete and the whole interface looks quite disorganized? Aren't the details of the downloading and the protocol communication supposed to be hidden from the user (after the algorithms have been studied and implemented as efficiently as possible)? I counted something like 142 options in aMule's preferences, and I am wondering, isn't a program supposed to do as many things as possible, automatically, without being needy or asking for orientation for every detail of its behavior?

Is the aMule team interested in a code cleanup increased automation and reconsidering the offered options?

Anyway, I guess I am trolling. Sorry! But thanks a lot for the answers. At least we agree that smart LowID check is useless.

[greek]

In the wiki, it says that I.C.H. takes only little CPU power, http://www.amule.org/wiki/index.php/ICH .
So if that is the only reason, it is not a good one. I.C.H. (or AICH) seems like an integral part which characterizes aMule and should be on at all times and do things transparently. So this option is a left over from the past and I propose it as a candidate for elimination, along with smart LowID check (and maybe that progressbar/percentage one).

[Kry]

Removing options is not part of this project's philosophy. What we do is provide reasonable defaults.

Smart lowid check is not redundant, or unuseful.

As for the rest, if you don't want or need to chagne those options, just don't change them. Don't force other people not to be able to change them just because you don't want to.

[lfroen]

Removing options is not part of this project's philosophy. What we do is provide reasonable defaults.

Philosophy has nothing to do with redundant options - just note.

I guess the reason for "smart low id check" being an options, is because it not always work.
ICH indeed takes little CPU %%, but on some systems (like embedded one), event this is too match. For example, on my NSLU box I have to turn off KAD, because of high memory usage.

[greek]

Removing options is not part of this project's philosophy. What we do is provide reasonable defaults.

Philosophy has nothing to do with redundant options - just note.

Thank you!!
So, lowid check is not always working because it's implementation is incomplete or because of occasional "rough" conditions?