@navi_wotd. Na'vi word of the day on Twitter.

[Kekerusey]

Hi,

Is it possible to get all these audio files in a single zip file or some such? I ask because I think it would be quite cool to simply put them all on an MP3 CD and play them in the car much as I once did with a similar archive of just Na'vi words a while back.

Keke

[Puvomun]

Kaltxì ma Keke,

Is it possible to get all these audio files in a single zip file or some such? I ask because I think it would be quite cool to simply put them all on an MP3 CD and play them in the car much as I once did with a similar archive of just Na'vi words a while back.

A zipfile, yes. A some such... I'd have to design that first.
There are 522 soundbytes at the moment, totalling 90MB. I can create the file for you and set it up for download on my kelutral server. Note that this would be a manual thing, not something that is automated (at the moment). If more people like this, I can set that up.

ta Puvomun

[Plumps]

I would also be interested in that 'cause I think it's one of the few regular listening exercises we have the moment

[Puvomun]

All right then.
When you go to http://kelutral.dyndns.org/download you'll see the ZIPfile link on the page.
Enjoy!

[Yawne Zize’ite]

For stress-marking, what about the English tradition of putting the stressed syllable in all caps?

KAme
PA'li
KELku
vospxìAY

[Puvomun]

For stress-marking, what about the English tradition of putting the stressed syllable in all caps?

KAme
PA'li
KELku
vospxìAY

That is a great idea. I'll try to think of that and apply it when I am certain of where a stress goes. And if I forget: no stress.

[Plumps]

Thanks for the .zip package
»Karma, vingardium leviosa!«

[Puvomun]

Thanks for the .zip package
»Karma, vingardium leviosa!«

Irayo si ngaru, slä nìprrte'

[Kekerusey]

A zipfile, yes. A some such... I'd have to design that first.
There are 522 soundbytes at the moment, totalling 90MB. I can create the file for you and set it up for download on my kelutral server. Note that this would be a manual thing, not something that is automated (at the moment). If more people like this, I can set that up.

Irayo ma Puvomun

Have a look in the Project NagNume sub forum  ... following a conversation with Tirea I've started a thread proposing NgayNume II and seems to me that resources like are just the sort of thing such a project needs.

Keke

[omängum fra'uti]

Irayo ma Puvomun

Have a look in the Project NagNume sub forum  ... following a conversation with Tirea I've started a thread proposing NgayNume II and seems to me that resources like are just the sort of thing such a project needs.

Keke

Before being used as a learning tool, it would probably be good to be reviewed by a few well trained aynari and aymikyun. I'm hardly infallible and there are plenty of mistakes in grammar, usage, and pronunciation. I'd estimate at least 10% have some form of error on my part.

Ma puvomun, if it would be easier I could add zip file creation to my upload process pretty trivially. I do occasionally go back and update older files with corrections/omissions (Haven't recently, but now that my time isn't taken up with wedding planning, I may have more chances to do it) and it would be easy for me to just archive the directory every time. I'm going to guess that your collection doesn't revisit anything older than current (As your quote of 522 sound bites is a bit low; as of today I have 546, some of which were days I missed or was a little late on so probably aren't in your list). Alternatively, I could produce a report of what has been updated that you could use as an index to download older files that are changed.

[Puvomun]

Ma puvomun, if it would be easier I could add zip file creation to my upload process pretty trivially. I do occasionally go back and update older files with corrections/omissions (Haven't recently, but now that my time isn't taken up with wedding planning, I may have more chances to do it) and it would be easy for me to just archive the directory every time. I'm going to guess that your collection doesn't revisit anything older than current. Alternatively, I could produce a report of what has been updated that you could use as an index to download older files that are changed.

Indeed, my script just pulls the MP3 that is needed for the current WotD and does not check back for updates (I don't even see a possibility for that at this moment as I pull the file in through wget over http).
We'd have to figure out a good way for transferring updated sound files, it would be a waste of time and bandwidth if I keep pulling in the previously changed ones time and again. An idea could be - if this is not too much work for you - to add a column to your spreadsheet in which you add the date when you updated a sound? I could code something to check that against the sysdate and pull in all the newer files since then. Just an idea, perhaps you have a better concept.

[omängum fra'uti]

Indeed, my script just pulls the MP3 that is needed for the current WotD and does not check back for updates (I don't even see a possibility for that at this moment as I pull the file in through wget over http).
We'd have to figure out a good way for transferring updated sound files, it would be a waste of time and bandwidth if I keep pulling in the previously changed ones time and again. An idea could be - if this is not too much work for you - to add a column to your spreadsheet in which you add the date when you updated a sound? I could code something to check that against the sysdate and pull in all the newer files since then. Just an idea, perhaps you have a better concept.

That could work. Most the time if I am fixing it, there is a slight change to the text as well. But not always. And it is an extra step that can be forgotten. (I'm a very forgetful sort.)

What I had in mind is a file in the same location as the sound files, which contains a log of modifications. Basically a text file with a filename or number per line or something like that, constantly growing. Lines could be repeated any number of times as they are updated, but just check new lines for what to update. (Alternatively, just one line per file, with a timestamp/sequence number/etc that gets updated when the file is modified.)

However there are two downsides there; one is that it is constantly growing (But then again so is the spreadsheet, and at a faster rate). The other is that it doesn't just catch past updated files, it catches future updates as well. For example, being that I was going to be away for two weeks, I had about 20 pre-recorded and uploaded, and all of those would have been listed.

[Puvomun]

Indeed, my script just pulls the MP3 that is needed for the current WotD and does not check back for updates (I don't even see a possibility for that at this moment as I pull the file in through wget over http).
We'd have to figure out a good way for transferring updated sound files, it would be a waste of time and bandwidth if I keep pulling in the previously changed ones time and again. An idea could be - if this is not too much work for you - to add a column to your spreadsheet in which you add the date when you updated a sound? I could code something to check that against the sysdate and pull in all the newer files since then. Just an idea, perhaps you have a better concept.

That could work. Most the time if I am fixing it, there is a slight change to the text as well. But not always. And it is an extra step that can be forgotten. (I'm a very forgetful sort.)

What I had in mind is a file in the same location as the sound files, which contains a log of modifications. Basically a text file with a filename or number per line or something like that, constantly growing. Lines could be repeated any number of times as they are updated, but just check new lines for what to update. (Alternatively, just one line per file, with a timestamp/sequence number/etc that gets updated when the file is modified.)

However there are two downsides there; one is that it is constantly growing (But then again so is the spreadsheet, and at a faster rate). The other is that it doesn't just catch past updated files, it catches future updates as well. For example, being that I was going to be away for two weeks, I had about 20 pre-recorded and uploaded, and all of those would have been listed.

That would work as well. It is not difficult to set the cut-off date to check to sysdate, so any update/change that is set in the future could easily be skipped. Can you drop me a version of your log so I can have a look?

[omängum fra'uti]

Right now I don't have a log, but it would be trivial to add in any format. The log wouldn't have date of the word being posted, just the date and time it was last modified, if I went that route. So you'd have to go off the word number being in the future.

For now if you wanted to do a one time manual refresh of old updated words, I just created a "cliplist.txt" (Same location) which has md5sums of all the files. I'm not sure if md5sum would be the best option for the future update process, but for now it's a good way to catch what is different.

[Puvomun]

Right now I don't have a log, but it would be trivial to add in any format. The log wouldn't have date of the word being posted, just the date and time it was last modified, if I went that route. So you'd have to go off the word number being in the future.

No problem either. I know the word number.

For now if you wanted to do a one time manual refresh of old updated words, I just created a "cliplist.txt" (Same location) which has md5sums of all the files. I'm not sure if md5sum would be the best option for the future update process, but for now it's a good way to catch what is different.

Great, thanks. I'll try to have a look this/my evening (I am 9 hours ahead of you), and otherwise I should be able to do something in the weekend.
We'll make this work.

[Puvomun]

For now if you wanted to do a one time manual refresh of old updated words, I just created a "cliplist.txt" (Same location) which has md5sums of all the files. I'm not sure if md5sum would be the best option for the future update process, but for now it's a good way to catch what is different.

Great, thanks. I'll try to have a look this/my evening (I am 9 hours ahead of you), and otherwise I should be able to do something in the weekend.
We'll make this work.

Wiya... I can't get to it just like that. Could you PM me the proper link please? Irayo.

Ok, got it. Typos rule.

[omängum fra'uti]

Excellent! You might want to upload the webpage with all the words to have links to sound clips that were missed before, as well.

I've updated my tools to append to "upload.log" (same location) every time sound file is put online. Each line is just a record of a single upload, recording the word number uploaded only. This file will just grow so just keep track of what line (or file offset) you last looked at and go from there. (And be prepared for the same sound clip to be listed multiple times since you last checked.)

As a test of picking up older clips, I also did an older word today and uploaded that.

[Puvomun]

Excellent! You might want to upload the webpage with all the words to have links to sound clips that were missed before, as well.

I've updated my tools to append to "upload.log" (same location) every time sound file is put online. Each line is just a record of a single upload, recording the word number uploaded only. This file will just grow so just keep track of what line (or file offset) you last looked at and go from there. (And be prepared for the same sound clip to be listed multiple times since you last checked.)

As a test of picking up older clips, I also did an older word today and uploaded that.

I started writing a program that runs a check on the number of soundfiles from your cliplist and what is present on my machine, as well as a check on the md5sum. The aim is that it will create a script that pulls in any missing files and also changed files. As the webpage is recreated every time, that will then also reflect possible missing soundfiles that are added later.
As I am not bathing in time and also writing this in C# which is not my strength, it may take a while. (writing it in python or php would take a while as well so I can just as well get used to C# in the process.)

[Puvomun]

Okay, I am having way too much fun with this. I created an md5sum file for the mp3s I have here. Using that and your cliplist.txt, my program now tells me the names of the missing files and the names of the changed files!  

/edit/: it's a good idea to do a rm of an mp3 before doing a new get. Just to avoid funny filenames.

[Puvomun]

A new words-file is available at the usual spot...