Sponsorship for specific bug-fixes

[ihug]

I work for an ISP in New Zealand, and we're (somewhat bravely) using RoundCube as our public-facing webmail service.

We've had lots of feedback from our customers, and while many see RoundCube as a great improvement on our old webmail system, a minority have had problems, and a few are so upset they're even threatening to close their accounts.

We have some in-house developers, but they lack experience in the product so they can't yet provide sufficiently rapid responses.

So we have a financial interest in seeing the project advanced as quickly as possible, and therefore are willing to offer sponsorship either to the project as a whole, or to individuals if needs be.


The issues we need immediate assistance with include:

• When deleting multiple messages, move them to the Trash folder rather than deleting outright.
  Preferably go shift-click-on-delete (or similar) to delete messages outright.
• Some types of in-line content are not displayed.
  I suspect they're construed as "attachments", but no link to them is presented.
• If one selects multiple messages then deselects all but one, it behaves as if multiple messages are selected and triggers the "really delete" prompt.
• Problems logging in with Netscape 7.2 and Safari on MacOSX, and possibly other browsers too.
• User-configurable session time-out.
• Sensible error message for "attachment too big"; sometimes attachments disappear between the user and the recipient, other times they just stop uploading without saying why.
• More memory-efficient handling of attachments, thus allowing larger attachments.
  (We have customers who want to send 20MiB attachments, which bloats Apache up to over 150MiB process memory.)
• Sometimes shows "no messages" at the end of a list of messages [we have a patch, where do we send it?]
  
• On a new message, can't add a recipient from the address book [solved - but needs documentation]
  

Obviously we'd be delighted if some of them are already fixed, or are simply a matter of configuration.

Which form of payment would people see as being more appropriate: sponsorship to the project, or payment to individuals? And what is the going rate? (We're expecting somewhere around 100-500 Euros per bug, but hourly rates would also be considered.)

Reply here, or by email to roundcube-maint@ihug.co.nz.

[Florianer]

• on a new message, can't add a recipient from the address book

Hi,
as far as I know, in composing a new message you cannot open the addressbook and see all the contacts, but it is possible to 'access' the contacts. Start typing email or a part of the full name / surname, a list of all contacts matching to this typed in part will appear. Hope, that helps, but I guess you would like to see the addressbook to browse trough all the contacts...

[kmn]

Hello,

It is possible to see all the contacts from the address book. You have to set the no. of records in the personal preferences higher depending on the number of contacts. By default it is 40.

kmn

[ihug]

[when] composing a new message you cannot open the addressbook [...] but it is possible to 'access' the contacts. Start typing email or a part of the full name / surname, a list of all contacts matching to this typed in part will appear.

That was pretty much what I was after, thank-you.

And to see "everyone", it suffices to search for "@", since every address has to have one of those.

[kmn]

Glad it helped you Martin.

But beware, this makes page loading slower. So your users who are on slower connections may start complaining about "slow mail". Tell them this is a trade-off between convenience and speed.

I have rolled out RC in a semi production environment (very bravely, as you have said, but crossing my fingers) on a beta basis. Around 500 users. Not many problems so far. But there are some really bad problems/lack of features with the existing SVN. I am sure your users are also going to come back to you with these problems.

1. Opening attachments is a real hassle. It is good to have (a) "download attchment" link near each attachment and (b) "download all attachments" for multiple attachments. There is a modification in the forums for (a) suggested by a user, but it does not work with the SVN. I think (b) has been accepted in the trac. Could you have a look at (a). The forum link is
http://roundcubeforum.net/forum/index.php?topic=780.msg3349#msg3349

2. If a user deletes his lone default identity, he is no more able to send messages. It is necessary to prevent users from deleting his lone identity. I have for the timebeing disabled the identitities tab for preventing this.

I see that you have mentioned you have some developers with you. Could you consider looking into some of these very essential modifications for the code, which you could contribute to the community.


kmn

[ihug]

But beware, this makes page loading slower. So your users who are on slower connections may start complaining about "slow mail". Tell them this is a trade-off between convenience and speed.

We don't plan to let them have large address books anyway.

I have rolled out RC in a semi production environment (very bravely, as you have said, but crossing my fingers) on a beta basis. Around 500 users.

How do I tell how many real users we have?

Code: [Select]

mysql> select count(distinct username) from users;
+--------------------------+
| count(distinct username) |
+--------------------------+
|          138884 |
+--------------------------+
1 row in set (2.92 sec)


Not many problems so far. But there are some really bad problems/lack of features with the existing SVN. I am sure your users are also going to come back to you with these problems.

1. Opening attachments is a real hassle.
2. If a user deletes his lone default identity, he is no more able to send messages.

We've had some users report problems with vanished attachments; some sort of fix would be useful.

I don't think most of our users are sophisticated enough to "fiddle" with anything labelled "settings", but no doubt someone will do it (and complain) soon.

I see that you have mentioned you have some developers with you. Could you consider looking into some of these very essential modifications for the code, which you could contribute to the community.

Neither I nor our other developers are yet familiar enough to debug things, let alone write patches. We have some good general-skills programmers, but things like Ajax are still foreign territory and so it'll take us a while to get up to speed.

If someone would be willing to spend a bit of time nursing us in the right direction, we'd very much appreciate it. We're in UTC+12 so an email conversation spaced out over days is rather frustrating; I don't suppose any senior developers would be willing to take a phone call?