Virustotal.com uses upwards of 40 different scanners to check a file. Granted, it has a 10Mb limit for file size, but I think a large majority of mods probably fall under that number. I use NortonIS at home, but I still find myself using Virustotal for any file I download small enough to send there. Never hurts to CYA.

As for other issues, especially nay-sayers, I would like to think that the offer to just link a user to mods on their collective websites, forgoing an auto downloader/installer, would be enough to resolve most issues or concerns they have. If it doesn't, well, I don't think anything we could offer would change their minds. In essence, it would be no different than what Yacoby's ES Search performs now. I don't recall seeing any objections to that.

So, is this something you're seriously considering running with, peachykeen? If so, I will give you any assistance I can, even if its just monotonous research and legwork. I recall a thread that had a list of modders giving their permissions for usage of their mods/intellectual properties. Is there a similar list for uploading permissions? As I've said before, I know a lot of modders prefer to control the release of their mods. Just wondering if we should find out how much (if any) interest there is out there for modders to actually use this centralized database. No use in doing a lot of work on your part if only two mods wind up on the server.

I would just like to point out that `normal` mods cannot contain viruses as there is nothing in text esm/p/s or images files that is executed.

That can be used to scan before upload, and an option added to use an installer virus scanner (if one exists).


I like the idea of having it install, but for those mods that can't be hosted, just linking would work (maybe with an install-from-local-archive function).
I was thinking last night, and I don't know how it works, but mlox already does a good job of scanning for known conflicts. A program that combines a simpler downloader with a Mash-style installer/manager and mlox conflict detection (without having to mess with Python) would be nice, I think. It would take translating the mlox code to another language, or wrapping a Python interpreter so it could be called within the program.


Yeah, I've been thinking about it. Last night I mocked up a GUI (pretty simple, but functional). I'm not sure at the moment whether to do it in C#/.Net or C++/CLR (which still takes .Net, I think). MySQL has connector libraries for both, so it would really depend on what other features would be needed and whether any of them restrict it. Any extra functions C# might offer can probably be done using Boost, however C# is faster as far as developing simple GUI apps (it's not quite a fast overall, but isn't slow).
I'll be better able to pick after I look at Yacoby's code for mods, but it's between those two ATM.

I don't know about permissions or whatnot yet, I think it'd be better to know the program will work (or can be made) before playing with all that. I had some thoughts as to a server, possibly a simple PHP admin function to allow modders to update their mods. That might help with some of the complaints.

Yep. Mlox is what got me thinking about this. I would also prefer that it auto-install all mods, but I guess there are too many hoops to jump through for that. I think what we've proposed here is a good compromise that should work for everyone involved, even those who don't want to let the program auto install directly from their sites. I like the idea of a Mash type installer/manager with mlox conflict detection. That might save a lot of work if their respective creators give the okay to use code from it. Although, as you've stated, Mash uses Python, which my computer really hates (must be age), so I think coding with C++ is a better idea. I believe OpenMW is also currently translating to C++. Someday I need to settle down and take a few more programming classes. I didn't like VB, but maybe C++ will be better. Of course, you're talking to a guy who originally learned programming in high school on a Macintosh a few decades ago.

Might be easier to tweak mlox so that it supports a console version, or some of the functions can be called from C++. That we don't end up maintaing two versions of the same program, saving work. There is Python for .NET or/and IronPython, which could be worth a look at for interfacing C# with Python code without re-wrtiting it.


I am not sure why your computer hates Python, I have it running on a computer that was in 2000. Be aware that there are disadvantages to C++, in that it is slower to develop with due to the lack of things like garbage collection and a small standard library. The speed advantage of C++ over C# is a lot smaller (in effect, nothing) when the program is I/O bound (which this will be due to downloading files off the internet, extracting archives etc). I would suggest C# or Python.

I cannot be involved in anything for the next 3 - 4 weeks due to real life commitments. I also cannot develop on windows (well, I can, but I won't ). C# is still viable due to Mono although I am not sure what GUI support there is. I seem to remember it uses GTK rather than Winforms. IDK.

Mlox could use a rewrite at this point anyway, as it has grown by the standard method of program accretion I might get around to it, but probably not real soon.


I've heard good things about Hans Boehm's garbage collector for C/C++ programs, if you want to add garbage collection. It's usually much more fun to program when you have GC.

I must admit I have never used it, although I have also heard good things about it. It does sort of strikes me as like nailing legs to a dog to try and make an octopus

Well, since mlox is due for a rewrite, as you put it, would there be any problems if peachykeen incorporated it into something like this? The two of you could figure out the best way to include any improvements you might have in mind, thus saving you from having to redo the program or maintain it separately. Of course, its such a nice program, an updated version of it would be nice to have outside of the downloader/installer for those who choose not to use it. I guess its always nice to get permissions first.

Well, an octodog can make a nice conversation piece



Not at all. mlox is under the MIT license, which means you can already do whatever you like with it. If there are any specific suggestions for what should go into the next version to make it easier to use in other programs let me know. But like I said, Bog only knows when I'll get around to it

Million things to reply to...

So, since my last post, I've mocked up a working copy of this. Actually, I installed MySQL x64 on Win7, it started complaining, tried using an old install of WAMP server, it wouldn't work, rebooted under XP, tried WAMP there, went and got the MySQL installer, installed, played with it for a few minutes, imported Yacoby's database, started programming in C++, the MySQL connector was being a pain, decided to scratch that, installed VC# Express, and made what I've got now. It was an eventful day.

Anyway, GUI shot:
http://i4.photobucket.com/albums/y145/peachykeen000/gui3.jpg
I'd upload a copy of the app, but unless you've got a MySQL server with the database available on localhost, it won't do you a lot of good at the moment.

It works with Yacoby's database, and can draw all the needed info from there (I had to learn JOIN for that). It runs reasonably fast, under a minute to load ~12000 mods from the server (it would be slower over the internet, though). Filters out just the MW mods (since that db has both MW and Obvs), adds them all to a list, then builds the table. Click a row in the table and it'll grab the info and show it down below (the which-row-is-which-mod logic isn't quite done yet, so it gets mixed up if you resort the table). Obviously, all the buttons and whatnot go down there. It shows the URL, blah blah blah.

As far as language, I've decided on C#. Faster programming for this, no real speed loss (speed is limited mostly by how fast Windows can draw the table), basically everything Yacoby mentioned. I don't know Python (I can make it work when I have to, but I'm not familiar with it), hence C#.
It uses .Net 3.5SP1 at the moment (it had linked to 4, but I told it to use 3.5 instead). I don't know what issues that'll cause with Mono/OSes other than Win, but I have a Ubuntu system I may be able to test that on (I'm playing with what's on there right now, hence my server being funny sometimes and my random avatar disappearing). I'll have to look at how Mono works (this uses Windows Forms, but no hacks or anything crazy, so it should be pretty easy to port). As a funny side note, I was playing with C#'s publisher function, and the first installer it made was 666 kb. I'm going to take that to mean powerful forces are smiling up upon this unholy endeavor.

As far as mlox goes, I'd rather bind it than rewrite it. It'll allow for much easier updates (just install mlox into a folder and point this towards it) and is just less work in general. I'm not sure the best way to interface them, I'll have to look at the .Net <-> Python stuff out there and how mlox works. I'm looking at Python for .Net and IronPython now, depends on which will work better for that. John/Yacoby, any suggestions there? If IronPython allows mlox to run within this, without needing Python in general, that'd help with the whole non-technical aspect of having one manager without all those requirements.

I'm looking around for various archive programs with .Net/C# APIs or interfaces or whatever... I think, for installing, it'll be far easier to extract the mod then install, instead of trying to install from the archive. I'm looking at a simple copy installer, but I have code to parse ESP files in C# already, so a complete intelligent installer (reads the mod, figures out what it needs, and puts that where it should go) is also possible.

I don't have any internet functions working at the moment, past the MySQL client. It can connect to any server to use the database, it just has to be in Yacoby's format of two tables and all (that's easy to change, though, just one query to retrieve the whole thing).

Last, but certainly not least, this needs a name. I don't have any ideas, so the test build is "MWModThingy." Any ideas?

I'm not familiar with any of these things, I've never done any real development on Windows. I do know that Python interfaces very easily with C, but that's about the extent of my Python knowledge

As Vader would say "Impressive"

I liked the table - wondered about a sort by mod type filter - in my repository I store mod collections by - type eg clothing, armour, weapons, quests, resources.

I like how you added the link to the original mod site so it's possible to go see the mod - and check if a screenshot exists.

Quite a few mods have readme's that state the mod is not to be uploaded to another site without their permission - how does that fit in with this? If you press install do you still have to go via the original dl site to get it?

How do you know if the mod is clean or not? Often comments on PES by other users give warnings if a mod has evil GMSTs or if it has other compatability issues

Well, I've download mlox and IronPython, I'm going to play with those tonight, see if I can get something working. Apparently Python interfaces with C# nicely, or so the site says.
If you don't mind (I know it's technically allowed, but still), I may make some modifications to make mlox work with this, until you have time to update it. Like a temporary app-specific copy, just until it can all be designed to work together.


I wonder how we could tie MW into that and make a name.


There is a "category" field, which doesn't have much at the moment. It's possible to add that to the table though (as a column), then you just click and it sorts alphabetically.


I'm planning (and this part is only in the thought stage) to have it check whether a local copy exists. If so, the "install" button will be enabled. If not, it'll just show the link. There will still be an installer, but if it can't be hosted, you'll have to do the downloading yourself. That offers a simple one-click mod download-and-install system for new users and quite a few mods, and lets it still be compatible with the rest.


The ESP-parser doing the installation could check cleanliness before enabling the ESP and ask if you'd like to clean it. Shouldn't be too hard to add a list of default GMSTs and run through.


I've been looking over Mono. I don't know how converting from Windows Forms to Mono would go. They say it's not so hard. I grabbed all the various files (IronPython, Mono, and a few others) and I'll be playing with them more tonight.

Edit: And as far as some of the features go, I'm adding two more flags to the db: advlt content and local copy. advlt mods (and it wouldn't be a guaranteed rating, just a marking if something was known or caught) could be shown in red in the list (or a pink background), with an option to filter them out. Local mods would be downloaded straight from the server (with another link given), non-local mods would have a download link provided.

I don't mind at all. If you like, I could even sign you up on the mlox googlecode project page as a co-maintainer. Then you could check your changes into subversion. But whatever works for you is okay with me.

Morrowind was released in the waning years of the "Web 1.0" if you will, so lots of the mods are either scattered across quaint personal websites or PES/TESNexus. At this point in time, as much as I hate having to click through mirrors and log into an account to use these sites, I feel that they have fulfilled their purpose well enough so far and I agree with you on this. Most people will stop hunting down mods, giving feedback, acknowledging the existence of their creators. They'll just mindlessly let the leech program handle updating for them, and will never really make much of an effort to motivate the creators or anything of that sort.

There's just something about the system that's been in use for the past 7-8 years which works perfectly. The personal community and social involvement in modding is key to this game's survival. If everybody just downloads their mods through a soulless program, I feel as if they're far less likely to stumble upon other interesting mods. This is a pessimist generalization, but regardless, I myself enjoy hunting down mods and taking time to hand-pick them.


That's harsh. Torchlight's quite moddable and stands quite well on its own merits too. Because there's no strong "core" community yet, with perhaps two decent fan sites in existence, I can kind of understand their need to unify and catalog most of the mods through TorchLeech. Still, I am quite surprised that the game's mod community has developed as quickly as it has. Plenty of utilities (aside from TorchLeech), documentation, etc. I regret to say that I myself am looking to do a non-linear, large-scale hub-based dungeon campaign for Torchlight and it's eating into my Morrowind time!

I like where this is headed so far. Here's a possible suggestion that might be able to be incorporated. I don't know about everyone else, but I have a huge folder/dvd collection of mods that I've already downloaded. Maybe the program can create a dedicated folder on each person's computer that their downloaded mods go into (or can be placed in there manually from dvds). When the program starts, it can scan this file as well as the server for available mods. This would prevent a person from having to redownload mods that they already have or have downloaded/installed using the progam. That way, if they wanted to start another game, they wouldn't have to redownload mods from sites that don't wish to be included in the program. Hmmm. That may sound like I'm trying to get around these sites, but I'm really not. How many times do you really download a mod, anyway? Most people have them stored/saved somewhere. Am I making sense today, or just ranting? Reeeeaaally bad sinus headache today.

I'm not sure if I agree with that. I think making some of the boring parts easier will help. Sure, there will be people who will just click download and never join in the convos here; but those are the same people who just click download and never join in as is. I don't think a tool like this will demotivate the people who enjoy/appreciate this community, and as nice as it would be, I don't know how much it could do to motive those who don't. Past combining a few existing tools into a simpler system and adding a download function for some mods, I don't think much will change. I think it's worth a try.


If you wanna talk about soul-less, there's plenty of Morrowind players who download their mods with IE. I mean, Internet Explorer doesn't just not have a soul, it goes out of it's way to shred yours.
And while it is true that people won't be wandering the webs to get their mods, that's both a good and bad thing. Given a list, especially like this (the copy of Yacoby's database I'm playing with has almost 12000 Morrowind mods), they might be more inclined to just sit there are scroll down the list, see what's there that looks fun. The program doesn't even work and I've already done that. It won't be all the mods in one place, or anything like that, it's just an attempt to make them easier to get and use. You'll still have to pick which ones you want, and links will be provided to the original page.




That's definitely going to happen, at least to some extent.
Since it's just dumb to download a mod you already have, it will grab the index off the server (if it can), then overlay your mods on that. I can slip scanning your mod collection in between those (except, it'll have to either go by archive names or they'll all have to be extracted, scanning through the archives would take a while). There could be an option to rescan the directory, or it could do it during idle time, but actually going through a thousand+ archives isn't a fast task.
I'm thinking of color-coding mods: green for you have it, but not installed; blue for installed and running; gray for installed but deactivated; red for advlt (and so purple for running, yellow for have it, dusty red for turned off). Something simple like that, so you can tell at a glance what's going on.
At the moment I'm working on turning the mod list into something that can be searched by title, which will allow for all of that to be done no problem.

Also, I'm working on making the link actually open something. Some genius at Microsoft decided not to put any easy to find shell functions in C#. As in, you search the docs for "shell execute" and get a "No Results!" In the Win SDK, you get a whole section. So I'm off to Google that up, cause otherwise you'll be stuck with IE to open the pages, and I'd rather it use the default browser.

Edit: Well, found ShellExec. Some idiot decided it was a diagnostic function. I swear, Microsoft needs to melt it's glacier once in a while and check out what the rest of the world is doing (and in more ways than stealing tabs from Firefox).

Hmmm, somebody's been busy. I like the color code idea. Heh. Someones not a Microsoft fan.

I have to admit that as I just read through the past two pages, consisting of many walls-o-text, I'm sure I've missed some things. I do want to throw out a couple thoughts I've had about this:

1. Local archive folder: When you first install the program you define the folder in which you have the mods you've downloaded. Then anything you put in there will be available to the installer. Or has this already been discussed, or a different method discussed?

2. Author feedback system: I think that one way to address one of the concerns mentioned here regarding lack of feedback for authors could be easily handled with some sort of internal system. There's nothing that prevents us from automating a system that sends email to the author with feedback. Or to even incorporate a comments database.

3. Submission standards: I am of the opinion that all mods submitted must be reviewed and approved, much in the way they are at PES. Of course this will require volunteers to handle, so is subject to discussion. Also, I see no difficulty in having server-side virus scans performed. I have also been working lately on a few scripts using john.moonsugar's tes3cmd to scan mods for things such as GMSTs and dirty records. I have been writing these for Morrowind Modding History, but fully intend to use them for this project as well.

Also, just a note regarding my involvement here. I have no intention in getting involved with the coding or development of this in any way beyond hosting and managing the server. I have way too many projects going on right now, many of which haven't even remotely been unveiled. I fully support this, I think it is a fantastic idea, and as such I offer my resources.

Since me last post here on this thread, this discussion might have some interesting arguments which I share and not share. At first I've hard to understand Skinjack's intentions in the first place, but now I do agree with you when mention the installation structure though. For my self I just hate to install mods manually, so your proposal about installer program you said earlier is a good idea. The only request a I've to this project is that, skip the feature of any web related components, just have it as a local mod-installer program instead. For those who still want access to the internet this might interest you is the 'Morrowind Mods Database', by Raveren55, as standalone downloader program. While I was searching for the correct link of http://www.search.yacoby.net/click.php?id=13877&location=NEX&url=http%3A%2F%2Fwww.tesnexus.com%2Fdownloads%2Ffile.php%3Fid%3D13877 I found the NOM-mod, I use ES Search to this. Finally I must say this, thank you very much Yacoby for ES Search, it's exellent tool to get information you want.

@ Fliggerty: Yeah, local archive folder. That's pretty much what I was suggesting two posts back in that rambling paragraph that you summed up in three nice little words. That'll teach me not to post when I can barely see straight. Stupid sinuses. Also, not trying to continually compare this to Torchleech, but it has a ratings column. Something like that could be included as well, so long as people didn't start handing out knee-jerk reactive "Zeroes" because they couldn't get a mod to work because of conflicts. I suppose it might also be possible to put an actual "Feedback" button that either links to the authors e-mail or a notepad of sorts, kind of like the comments on PES, etc. And I have to agree with submissive standards. Obviously we want mods that play well with others or DON'T TRAP PEOPLE IN BUILDINGS BECAUSE THE AUTHOR THINKS THE PLAYER IS CHEATING. Some of us just like to collect weapons and stick them in displays. Yep, still creased about that years later.

@C.Carjades: Well, its nice to see that we can agree, at least in part, on some of this. I'm still not sure why there is resistance to the web based downloading aspect of this, though. You already do exactly the same thing using IE, FIrefox, ES Search, but this way its all incorporated in one place. And the way we are trying to address sites such as Darknut's, which is dependent on ad revenue, is to simply direct you to that website to download the mod, which can be stored in the above mentioned local archive folder. Thus sites like his that have supported the community won't disappear. Actually, it might see more traffic than before.

I don't have any huge problems with M$ (save Vista), it just bothers me when they hide simple functions. I was able to get the shell exec working though, so the little links will open in your default browser when you click them.

I was also able to filter the list by name, so no matter which way it's sorted, it can still look-up the right info (it was way easier than I thought, just had to figure out how predicates and lambda functions worked and it took one line of code). I noticed an issue with that, though: if there are multiple mods with the same name, it'll only return the info for the first one, no matter which you select. I checked the database, and the mod IDs are all the same for those as well. Either I'm going to have to add an internal ID field (just for sorting purposes) or (I'm playing with this now) pick the one with the highest version.



The plan, at the moment, is to download the database, then scan your mod archive, then your Morrowind installation. All database mods will be shown plus archive mods, with various colors representing where they are.


That shouldn't be too awfully hard to pull off, but I'm going to get the from-web part done before working on any to-web stuff. I like the idea of a comment area, I was thinking last night about adding a rating field (but that has so many issues, I don't know). If I do add a rating field, it'll be off by default and you'll have to turn it on to see.


Well, a lot of the mods could be downloaded from PES, which has some screening already. It would be too hard to go through the mods, especially not if it was properly automated. When I start in on the install module of this, separating that code into a standalone extract-and-clean program should be simple.


Coding and dev I can take care of (unless of course you want to add crazy web stuff).
You say you'll be involved in the hosting, will you be hosting the final version, or what do you mean by that?
Because at this point I have a demo app that can retrieve the list and all, and I'm either going to sit down today with my server and router and make sure it's working, or if you'd rather use a server you have... Just need to get that all figured out.



You will be able to skip all the web stuff by not downloading the mod list from the server. In that case you can just sync with your local archive, and install from there. That's basically already built into the system, and will allow people to cut this off from the web entirely (if you really want to be paranoid, you can block it with your firewall).



I can understand the hesitation of having a program go out to the internet, find a file, and install it to your computer all in one click.
However. Everything will be scanned before being uploaded, there will be an option to scan on download (before install), and mods don't usually contain viruses, and can't run them (a virus file is harmless unless it's run, and within an archive, it's essentially encrypted (as far as the virus code is concerned), and can't run at all).
There will be no leeching from existing sites, just linking when necessary (and I think Dark0ne runs Nexus, not Darknut).
And as I mentioned, you can always firewall this in (most firewalls have a block list). That'll limit its functionality, but then you know nothing bad can happen.
And the source will be open, so if you can understand C# you'll be able to review it yourself (it's not hard to comprehend what's going on just by glancing through, much of it is English words and I don't think my coding style is too hacker-ish ).

I don't know what the other members would say about that, but I can maybe put some sense into that. Personally I belive this forum is very similar to the Linux community, where the user is allowed to develope / modify any Linux-apps they want or gain access to the open source program which is already hosted on many web sites today. My conclusion is that, both new or veteran member value the freedom to choose when, where and how to obtain mods for the own interests. That's something other member must respect and stick to.

peachykeen - Please, take a look on my previous post about the http://www.search.yacoby.net/click.php?id=13877&location=NEX&url=http%3A%2F%2Fwww.tesnexus.com%2Fdownloads%2Ffile.php%3Fid%3D13877 link I found on ES Search. Download it for a closer look, I might interest you. That would something to work on don't you think ? :talk:

This seems to be morphing into a fairly reasonable proposition .... making it all local is the best bet IMHO ...

I still think there will be a lot of issues trying to automate something like this .. for one thing a lot of mods have multiple esps, textures & mesh options buried inside the archive & there isn't anything even close to a "standard" for mod packaging.

I for one will prob never use it because I like manually installing everything myself & & if it is doe have web access I'm way to paranoid to let any program have that much access to my system.

Just My 2 Cents