If you want a yes or no from me, I would like to see all non-stealth applications banned as well, not because they are not portable but there are thousands of them in the Internet. If you really want them all on your web site you have to add a lot more to your database I guess. And in my opinion you have already too much repertory. Why not keeping it simple and take care of only the portable AND stealth applications? It saves time for you as well.
If you don't want here is another suggestion. In the feed list mark all non-stealth apps with an ! before the name.


That's why I would not differ between writing for reasons of settings or something else. Either an app leaves generally avoidable footprints at the host or not.


Maybe you should open another thread for this. But again, why not limit it only to the best of the bunch, so no edge cases are allowed at all? It sounds very hard but I believe there are heaps of applications left that fit. And these ones are highly portable, nearly stealth, genuine freeware and hence highly usable for everybody on every Windows system.

After sitting along the sidelines and reading the excellent exchange, I believe that I can summarize my preferences, now, as follows:

1. I prefer Freeware. Not because I am a cheapskate but, I cannot always determine whether a program is a "keeper" within the first 15 to 30 days that I have it. Those programs that I come to rely upon I do register.

2. I prefer Stealth. More out of an environmentalist approach in that, I don't like to modify the machines I am working on, any more than is necessary. Why should I clutter up someone else's Registry with remnants of my programs that I never installed? In my view, Stealth implies Portability, and is its highest form.

3. If I cannot have Stealth, I prefer Portability.

4. If I cannot have Portability, I carry the installation files. For example, I carry all of the installation and signature files for several Anti-Virus and Anti-Spyware products and install them when needed.

The way that the licensing is identified now seems OK to me. Perhaps all we need is a scale from zero to X, with zero being Not Portable (Rejected) to X, meaning Stealth. I would leave the definition of the scale to the TPFC Moderators.

Regards,
Dan

ok, i HAD a post, but, forget it,
i Really dont want to get into it,

all im going to say is:

A) Its Andys site and he can list whatever he wants, the site has been going along just fine with him doing that so far.

B) If you REALLY want to dictate whats in the database that badly, start your own site.

C) The word 'Stealth' or Non-Stealth in green/red next to the apps name or under the discription might help people looking for stealth only apps locate such programs without them having to read the 'writes to' section.
(though i dont think its that hard to do, )

And, if you think Non 'Stealth' apps shouldnt be listed here at all, see A and B,




Anyways.
i love Stealth apps, but i still use (more) semi-portable ones,
and JPE is nice for those that i want to make 'stealth',

(And usdcs, im posting under you, but, im coming off kinda ...blunt, and i dont want you to think your post did it,
im replying to the thread as a whole, after just skimming over it, )

So, yeah, thats it, my view on the whole thing in the list,

_________________
"It's not that I'm Evil, just that I'm unwillingly Good."
-Stolen, but oh so fitting. ^-^

Note: It might take me a while to reply to any post or comments due to offline reasons.
So if I don't reply or miss a question, this is why.

I think usdcs has summarised it very well, couldn't have said it better.

This thread has turned, in part, to some sort of ideology dispute, which is sad. I mean - we all love this site, and its popularity proves that Andy (and his tireless helpers) are doing a great job!

Let's keep it up as it is and let the users take informed decisions, which app they want to put on their device. A ranking scale might be an idea to help with this, although I'm afraid it's a large workload, and could spark off ideology disputes every now and then ...

I have found so many useful apps here, just want to say thank you once again.

Chris

I hope you guys don't get the word "dispute" because of my post. My post was an effort to help Andrew to get ideas on how to solve the complains regarding portable definitions and other things.
I assume Andrew is trying to set category so every people with different "ideologies" or POV would agree with that category or definition.
In my last post, I also included things that I don't agree but because some people think otherwise I included it also. I did give some comments or opinions with intention to help Andrew decides, but mostly I'm trying to be general or moderate about this matter.
This all because of complains, and complain because people have different perspectives of portability; if we have clear standardization of category, no one would complain.

I think one very easy way to give valuable information could be based on the proposal of Dan, slightly modified:

Red = "Not Accepted as Portable"
Orange = "Accepted as Portable" That is, that it uses the Registry but, it doesn't overly affect the usage.
Yellow = Portable. It might touch the Registry or write in the aplication folder, in the external support, but, is definitely portable.
Green = Stealth. No traces left on the host machine at all neither on the external support. It is not only 'portable' but also 'ROMSOFT'



*JT.

_________________
"Lo que tengas que hacer, hazlo pronto" (Juan 13, 27) / "What you do, do quickly" (John 13, 27)
http://es.wikipedia.org/wiki/Teobromina / http://en.wikipedia.org/wiki/Theobromine

Not at all, Chris!

I agree. I actually, started to write the additional comments below, in my last post. It seems that the discussion has gone full circle. (Maybe more than once!)


I believe that we've talked this through, and all that remains is write the definitions and to post them.

As an extension of that process, it would maximize my convenience if we also had a "portability rating". The 28 pound Compaq Portable was, by definition, portable. But, not nearly as portable as today's notebooks. In my mind, the "portability rating" is analogous to the PC manufacturers publishing the weight of their portable machines.

Regards,
Dan

[EDIT: Corrected spelling. Added weight.]

Ok, let's make it even more complicated. We have talked enough about registry and ini/folder settings, so far so good.

But in order to be really portable, imho a program should also systematically make use of relative paths, or - even better - offer an environmental variable to point to the root of the device it has been started from.

A bad example to cite here would be Save2FTP. Although portable with regard to settings and registry, it offers no possibility to enter relative paths, and is therefor only of limited worth for use from a USB-stick.

A positive example is Qsel, whose author wasn't initially aware of the implications of use from a USB-stick. However he immediately made it functional for relative paths (including a root variable) once this had been suggested.

This shows that there is more to portability than just registry and ini-settings. Admittedly, this is not of the same importance for all programs, for some it even doesn't matter at all.

Furthermore, many programs haven't been programmed with true portability in mind, and their portable versions are just some kind of variant, mainly with the registry in view. Might be worth to add the info about relative path support on the main page of the DB, where applicable (in the case of Save2FTP you will only find out about it when checking the comments).

Since we strive to convince authors to write really and fully portable programs, me deems that this point should not be forgotten.

What do you think?
Chris

Since windows keeps logs of what is run, they all leave traces - it is slightly advanced to try and delete traces of having been there.

To be portable the program should be able to run from that folder and work, ie load its settings etc from there. So you can just move the folder to another computer. If you can do that there is really no point of writing to the registry.

Appreciate all your comments. Thanks!

First off, I just want to let you understand where I am coming from. As I have mentioned many times in this forum and elsewhere, I started this site for two main reasons:

- I had a ton of portable apps in my possession, and I wanted a simple way to search them. This is still my main use for this site. Just the other day, I wanted to find out which app will let me wipe my SD card clean, and a few keywords later, I got my answer.

- I wanted to learn PHP.

As such, I have no ideological underpinnings, just pragmatic concerns. My usage pattern does not involve hopping from public machines to public machines, logging in as guests. I carry all my stuff on a 2.5" HDD, and I travel between my home and office machines. Sometimes, I bring the HDD over to a friend's place to help him troubleshoot. I also help maintain my wife's machine, and being able to copy apps over with all its settings (eg. Firefox Portable + all the extensions I have configured) is very convenient. Also, when I have to change machine for one reason or another (breakdown, upgrade), having portable apps shortens the downtime tremendously.

If I start sticking to some kind of purist regime, this site ceases to be useful to ME. I am not interested in sticking every PC app in the database. For example, a word processor that writes settings to the registry will never make it there, since I am tired for reconfiguring the settings. But if every data recovery or diagnostic app that touches the registry is not in the database, what use is it to ME? If on occasion I try to find an app that recovers the password of a Word document, but I can't find it in the database, what use is it to ME? And if this site ceases to be useful to ME, what motivation do I have to keep doing this?

I am suggesting what I hope would be a good compromise, and something that is simple to implement. I will add a boolean field to the database that says "Stealth: Yes/No", for those who don't want to read the "Writes settings to" prose. But it will be an optional field, like the "Unicode support" field. If the moderators have all the information to populate that field, then it will appear in the listing, otherwise not.

Lemme know what you think. Thanks!

Hello Andrew.

I find your website enormously useful and check it several times a day (yikes ... I need a life!). In fact, I have stopped checking other portable apps sites as your site gets updated rapidly, is fairly comprehensive, and gives me the additional information I need to make an informed decision about whether I want to try a particular piece of software or not. TPFC has become my one-stop shopping site for portable freeware.

Please be encouraged to continue doing exactly what you are doing now. The very fact that this thread has prompted some vigorous discussion is indicative of the tremendous value your site provides. If it was not as good and useful as it is, there would be no debate because no-one would care. But we do care, hence the passionate debate. Further, the debate, even when a little heated, always contributes to my understanding of the issues and I always learn from it.

The binary stealth/no-stealth field will be a useful addition to the information already given.

Thanks again!
/opsimathic

_________________
Better late than never.

Hey, Andrew: I see that I have the same goals (and obviously less knowledge) than you. In fact my 'way' to maintain my portable apps, was my webpage, which has a list of the app that I use regularly. Of course I would like to be able to build and maintain a page like TPFC... but I do not need it since TPFC is already on.

When I realized that I liked the apps to be portable, my only test was to install the program, to copy its folder into an external device and try to run it in another computer. Nothing related to registers, settings, etc. Of course the more I know about all that matters the more skill and feeling of domain I have of the portable apps... All thanks to pages like this one.


I have the same oppinion of opsimathic.

About the discussion we had, when my computer is full of programs or settings that I do not want anymore, I restore a 'clean' installation copy of my Windows OS by means of any restoration program -5 minutes- and I start again (my typical installations are composed of a 'clean' OS plus my portable collection). The main utility or purpose that I find for the portable apps is to start again my work with all my favorite programs in any place in a record time, as you do. But I know that we have to take into consideration some other situations like the use of the school computer, etc. For this reason is good to orientate the developers to have in mind the goal to create 'clean' portable apps, instead of apps that let traces everywhere.

Regards.

*JT.

_________________
"Lo que tengas que hacer, hazlo pronto" (Juan 13, 27) / "What you do, do quickly" (John 13, 27)
http://es.wikipedia.org/wiki/Teobromina / http://en.wikipedia.org/wiki/Theobromine

In reply to cmmehl's point about relative paths, I agree to a certain extent. For example, portable program launchers are pretty much useless without support for relative paths. Same with music players. As far as I can, I do try to point it out in the synopsis, and the comment area is always a great place for others to pick up my slack.

But for certain apps, relative paths are probably less important as a criteria for portability. For example, if a PDF extraction app remembers the path of the last PDF file it processes, it is probably not terribly important that that path is relative.

A gray area would be apps like OpenOffice that keeps track of recent documents you open in absolute paths. Although I have never been irked too much by this (I just File, Open and load the document again), I am sure some ppl will be hopping mad over this. Still, I don't think this behaviour should make eg. OpenOffice non-portable.

For me it's not the question, whether but where they do it. If they write the MRU to the host, they are not stealth, if they use a file in the app folder they are. Either way they are portable.
Or did you mean, they are arguable not portable, because you can't USE the MRU regardless of the method of saving? In my opinion this should not be a portable criteria, because it's inevitable until you remove this function completely. The same matter is with file association AS LONG the application does not catch them automatically or nerves in any tiny way to get them. A good example for this is IrfanView. A bad one would be an application like Winamp, which has an option to automatically cath them at every start AND where this option is switched on by default. It can be a little desaster when you use this type of apps on a friends computer.

For the rest of the discussion: I see my suggestions as suggestions and not as dictating. Peroid.
Andrew created this thread as an invitation to DISCUSS and that's what I do. That includes stating my point of view, my opinion and/or my desires. The same did Andrew and BECAUSE it's his site it's up to him to use the comments finally but it's not an initial argument. Anyway I often try to accommodate the most people (like Andrew). Of course this doesn't work always.

I think the thread reveals a general desire towards grading. Maybe it's just human to try to classify everything, to sort out the chaos. Obviously Andrew and a lot more don't care about this too much. I also would like to see a Wikipedia like site like this but with strict and easy verifiable rules regarding classification and SECONDARYLY detailed descriptions how to install or fixing the settings towards portability or stealth but until then Andrew's site is not that bad. In fact, I use it very often and I'm happy and thankfully this site exists at all.

To clarify, for example, if you have been working on a document called f:\docs\work.doc, so this goes into Writer's MRU.

Now you move to another machine. Your USB drive gets assigned the letter "g". When you use the "Recent documents" to try to load "work.doc", you will find it doesn't work because it is hard-coded to "f".