Roboform export experience with 1Password's conversion tool
Comments
-
I, too, am a long time user of RoboForm and was a total convert. But recently I've become very disillusioned with Roboform for two major reasons.
1. RF doesn't offer the possibility to sync my mobile devices locally. (most PW Mgrs don't). This is a must feature to me. I will sync my password database to the cloud over my (almost) dead body.
2. There is no way I've found to export the RF passcard data in any form that is usable by any other product or database. Even the 'Print all...' function munged the data so you can't easily edit out the logins and passwords. Once you have your data in their app, they've effectively locked you into their product. I was trying to export the data so I could merge it with another PW Mgr that did have local sync to mobile devices, but almost no browser support.
MRC;s tool is extremely capable, and a great tool. But, for me, it was not able to allow me to import my RF data. I tried to use MRC's conversion tool on my data that had been updated to RoboForm version 7.9. Even when I moved the data back to RF 6.9, the imported passcards did not work in 1Password. The only option successful for me is to recreate ALL the passcards.The way that 1Password handles the various browser logins is not nearly as seamless and reliable as RF, but because of the local sync feature, I'm really trying to make it work.
So depending on what are your must features, 1Password may or may not work for you.
0 -
I'd like to get some clarification here. @DoctorBrown, you've mentioned a few times about your converted data, imported into 1Password, that "the imported passcards did not work in 1Password".
There are two distinct pieces here being conflated, and they should be called out separately:
- The data conversion and the creation of a good 1PIF (that can be imported successfully into 1Password)
- The newly imported login records do not fill form data as you expect
It is converter's and my interest to be sure that (1) works. So if you, or anyone, has issues with this aspect, I'd like to know about them and I'll do what I can to correct any issues or shortcomings.
Regarding (2), a triplet of a username, password and URL imported via a generated 1PIF may be insufficient to support (2). In some cases it works fine, in other cases, the Login record must be updated so that 1Password can create the proper form-fill fields it needs to enter the credentials.
I just added Print Record export support for RoboForm for Mac. I'll look at version 7.x RoboForm imports, but as we know, the URL is abbreviated. Still, its better than nothing.
0 -
I'm sure I speak for many here that we are grateful for all the hard work you've put in to get the conversion to be as full featured as it is. It is a great tool and appears to be able to perform many types of conversions. It just that Roboform doesn't provide enough or the correct data to be able to be more successful at creating passcards that can successfully login to the target web site.
I can agree that I was not completely clear in some of my statements about importing the passcards and them 'not working'. I am also hobbled in my testing because I am running 1Password in Demo mode and it's not clear to me what limits Demo mode puts on creating, editing and using a large number of passcards.
Regarding the two distinct pieces in more detail:
- The data conversion and the creation of a good 1PIF (that can be imported successfully into 1Password)
I used the 'Print' function in RF to print 18 passcards representative of different types of web sites, to stay within what I think is the Demo limit of 20 passcards. After the conversion with the tool and import into 1Password, there was a passcard created for all of the entries. So, from the tool's and your perspective the import was successful.
But...
- The newly imported login records do not fill form data as you expect
Out of the 18 created imported passcards, for only three of them was 1Password able to: a) navigate to the correct web page; b) fill in and submit the form; c) successfully complete the login. That's only a 16% success rate. In that context the conversion was only partially successful.
For the passcards that can't successfully login, the process for editing a passcard to make it work and creating a new one is about the same amount of work. From a user's perspective, dealing a passcard that did not import is better than one that does not successfully login because of the work needed to find and fix the 84% that were imported but don't successfully login. For me, the conversion works for a given passcard only when it can create an entry that is able to successfully login to the desired web site.
I have looked at the raw HTML data as text in a HTML editor that is generated by Roboform 7.9. Unless I'm mistaken, it's going to be a b***h to get at the username and passwords. RF is inserting some kind of HTML or binary mark every third character or so in the username and password that might make parsing out the data pretty difficult, I think. In RF 6.9 the HTML output had the HTML field name which is required to fill in the form, but in 7.9 it looks like they are no longer always outputting the field name.
If the complete URL isn't available, the imported passcard has no chance to get to the correct web site and if the passcard doesn't have the data to fill in the proper field, that won't work either. For me it is not better than nothing because I have to essentially go through the process of creating a 'new' passcard.
I would be interested in hearing what your findings are when you look at the data in detail.
0 -
For 7.x, again, I'm going to stress the well-known issue... the issue is not with the conversion, or the import, but with the exported roboform print list data that does not represent the actual data within RoboForm. It is not faithful in exporting the required sign-in URL, but instead, outputs a top-level-domain URL. So 1Password may fail to auto-fill the data into the forms, because in many cases, RoboForm gave the wrong URL. Garbage In, Garbage Out.
So users will have to fix any abbreviated URLs within 1Password by navigating to the site's login area, and using the procedure for updating the Login entry. This isn't a big deal, and users should just do this over time as and when needed. The primary goal is to get the data out of RoboForm and into 1Password as quickly and efficiently as possible, so that users can move forward.
As far as parsing the munged data in the exported print list, that is trivial, and I already do so in the Mac and Windows 6.99 conversion.
I'll work on the 7.x converter this week, maybe today if I have the time.
0 -
I totally get it that the root of the problem is the outputted Roboform data does not contain the required sign-in URL, or other data needed to create a 1Password entry that will be able to successfully login. You are doing the best you can given the data that you are getting from Roboform. I also hear you that your goal is to get passcards in 1Password that map to the passcards that existed in Roboform. That is a reasonable goal.
My goal would be to create 1Password entries that are able to successfully login as close as possible as it would when run from RoboForm given the differences in functionality between the two PW Mgrs. In the best of circumstances, I know this is not likely to be 100%. For me, when I convert any data, contacts list, email, etc, I want to be confident that a high percentage of the data imported into the new environment is functional and correct. What would a user's reaction be if only 16% of the phone numbers and email addresses imported into a new contact list were correct? I'd forget about the conversion and just start doing a cut and paste from the old to the new for all of the entries I wanted to keep. (In RF 7.9 'Print List..' HTML output, even that is not possible. Try it)
Again, I have the same amount of work to do if I try to correct the failing logins or if I create them from scratch. It's actually easier to create them from scratch because 1Password prompts to save a new login if it finds it doesn't already exist.
What you are doing to work on a conversion is valuable and many people will love it. It's just that I have a different expectation of what I want from a conversion of this type.
0 -
2020autoglass: I apologize that MrC and I have hijacked your thread. Is there anything else that you would like to know about 1Password or Roboform?
0 -
Hi guys,
@DoctorBrown / @MrC, in the interest of keeping the original thread on topic and we can help the OP directly, I've split the rest of the posts into its own thread. I'll leave a link in the original thread if 2020autoglass wants to follow up here.
0 -
FYI, just in case you care, I've completed the RoboForm 7.x converter code. See the comments in this post.
0 -
@DoctorBrown and others,
Version 1.08 of the converter suite now contains the roboform converter that supports RoboForm for OS X and versions 6 and 7.x of RoboForm for Windows.
0
