MozillaZine

The future of Seamonkey?

Discussion of general topics about Seamonkey
tonymec

User avatar
 
Posts: 717
Joined: October 15th, 2004, 2:58 am
Location: Ixelles (Brussels Capital Region, Belgium)

Post Posted December 14th, 2015, 5:13 am

Ztruker wrote:If you had to leave SeaMonkey, what browser and mail programs would you use?

Good question.

I came to SeaMonkey because the Firefox developers were (already then) not listening to what users wanted, and removing features "behind the users' backs". The trigger, for me at that time, was the removal of the throbber link (remember, when you clicked the Netscape "N" throbber, and later the Firefox throbber, you arrived at some page characteristic of the browser). The answer, when I asked why they had removed it, was "nobody used it, and it wasn't discoverable anyway"; Oh yeah? So how come I (and several other complainants in the newsgroups) had discovered it and were using it? SeaMonkey was based on approximately the same code base as Firefox, it had kept the Suite concept that I knew from Netscape, it had not removed the throbber link, it could use themes based on the same principles (it even came with a so-called "Modern" theme in addition to just the default), it had similar extensions (even though, at the time, there was no "proper" extensions manager), and it had kept the "no-nonsense" Netscape-like Preferences dialog while Firefox had a very simplified Options dialog which, in comparison, I found barely good for kindergarten kiddies (with big pictures and few options). The "real" extensions manager came at SeaMonkey 2 (with the Toolkit Transition) and the Suite concept and the "serious" preferences dialog are still there (though with a different Preferences backend which changes nothing to the look and feel of the frontend). And of course there were yet other goodies, an important one of which (for me) was the built-in ChatZilla, which made me discover IRC chat.

Now what if the Firefox guys finally succeeded to throttle the lifebreath out of SeaMonkey? I might have a look at what Opera has become (there's been a long time I haven't) but I think I'll use Konqueror (and try to find a way to make it use its KHTML rendering engine by default, rather than Webkit). Too bad it isn't cross-platform but I have to deal with what I have. For chat I'll use Konversation, or maybe ChatZilla-in-Firefox (though it sounds to me like bloat to load the whole of Firefox just to use chat), and for mail… I don't know yet. Maybe go on with (blerh) webmail.
Best regards,
Tony

Redbugdave
 
Posts: 73
Joined: January 31st, 2012, 8:29 am

Post Posted December 14th, 2015, 5:49 am

Ztruker wrote:If you had to leave SeaMonkey, what browser and mail programs would you use?
I


I am in the same boat as you guys. I will stick with the Seamonkey ship to the last. As a last resort I guess probably Pale Moon if it is still around. I have used Netscape and then Seamonkey since they first evolved.

Frank Lion

User avatar
 
Posts: 20059
Joined: April 23rd, 2004, 6:59 pm
Location: ... The Exorcist....United Kingdom

Post Posted December 14th, 2015, 10:48 am

SM will be fine if Mozilla cuts them free, as they have just done for T/Bird, and I suspect they will. That will mean that SM doesn't have to follow Firefox down the non-XUL drain and can retain all of its extensions, themes and functionality and move forwards.

Will there be hurdles after that? Sure, the same ones that Firefox faced when it started and they managed it.

Free/low cost servers/hosting, etc is everywhere in the Open Source community and the SM Council just have to man up and make a plan. Code freeze for a year, concentrate on security updates and tidy the old girl up a bit. Don't forget that the extensions and themes to turn SM into Firefox as it was years back, already exist and add less that 500kb to the total.

Don't start reaching for your inhalers and talking about grippies and stuff to me, no one said anything about removing anything. But, to succeed, or even exist in the future, SM will need to look and work like a browser in 2015 should. When all is in place, then hit the media and get the userbase figures far higher. With them comes search engine revenue and at that point SM's future is assured and you can relax.

Finally, what's all this talk of other browsers? People endlessly describe themselves as 'loyal users' and yet are planning to jump ship at the first sign of trouble? What sort of loyalty is that, you can actually help SM in the way they have helped you over the years, you know.

Even pure self interest should tell you that. Vivaldi? - yeah, they talk a good browser, but in reality your RAM use will be over 4 times what it is now. It is pig ugly, limited functionality and nowhere near as customisable as the hype on it makes out. Go test it, like I have already done.
Metal Lion latest SeaMonkey & Thunderbird Themes - Sea Monkey and Silver Sea Monkey
"The only thing necessary for the triumph of evil, is for good men to do nothing." - Edmund Burke (attrib.)

rsx11m
Moderator
 
Posts: 14425
Joined: May 3rd, 2007, 7:40 am
Location: US

Post Posted December 14th, 2015, 12:59 pm

Yeah, I don't understand the overly pessimistic mood here either; the SeaMonkey ship is catching on water (we've been there before, haven't we?) but isn't quite a Titanic yet!

There is a corresponding sticky for Thunderbird at viewtopic.php?f=28&t=2976617 if you want to follow that discussion in parallel, and a collection of links to newsgroups and mailing lists was provided in the last status meeting (2nd paragraph of that section) for further reading.

So, what's the situation? Mozilla has made it unambiguously clear that they want to kick Thunderbird (and implicitly other projects like SeaMonkey and Lightning, without those having been mentioned - yet) off any infrastructure that's shared with Firefox, and that they won't do any "favors" for comm-central applications in the future any more (like establishing version branches on release repositories, etc.). There are two issues here, one being the actual infrastructure for code management and building, the other the code base, including the XUL/XBL/XPCOM/etc. code that both SeaMonkey and Thunderbird currently heavily depend on.

Cloning a repository to Github or Bitbucket is trivial these days, as Frank already pointed out, having some decent build machines in place (other than in someone's private basement) is a different question. But, all of this would be "somehow" solvable, and it is not clear if and when Mozilla kicks everybody off their servers for good.

The code base is the more burning issue, with a couple of options having been / still are under consideration, either of which is unsatisfactory and associated with substantial work:

  1. Following down the slippery slope Firefox is pushing for, with a further degrading (feature-wise) Gecko, thus coping with the removal of key features and essentially turning everything into a huge web app on a local minimalistic browser backend. This is apparently the way Thunderbird wants to go within the Gecko 45 and 52 ESR lines (if the latter will still exist), then having to look for alternatives like Servo. For SeaMonkey, this variant would introduce some painful changes in the well-known user interfaces (e.g., in-content preferences, which in turn could be rendered in a separately opened browser window without any toolbars, workaround like that as possible). On the upside, SeaMonkey and Thunderbird could work on the transition with combined forces, despite differences in UI, but adaptions to the shared MailNews code would be of common interest. Obviously, some features may not be implementable, that will probably have to be figured out as we go along.

  2. The other extreme is to get off the Mozilla train and coordinate with other interested groups a continuation of Gecko as is, maintaining its configurability and key technologies. While this sounds "nice" to start with, problems here are regarding security updates and developing web standards. Meaning, such a solution can't be "static" and someone would be needed to understand security implications and to continue a secure code base. Unless Thunderbird follows that path, this would also imply backporting any MailNews core changes, which may become difficult if and when Thunderbird switches to Servo or whatever other option they choose. PaleMoon has been mentioned in this thread already, and it might be an option for collaboration. However, they forked off the Gecko 24 ESR branch rather than the upcoming 45 branch, thus a lot of API changes made by Mozilla during the last two years would have to be reversed while not touching any improvements in either Core or MailNews code. Also, it is unclear what would happen if SeaMonkey wants a specific feature while the main PaleMoon developers don't want to introduce it into their Goanna rendering engine.

  3. Another option suggested was to run SeaMonkey itself with a possibly stale Gecko version while somehow embedding a current rendering engine (and possibly Servo once ready) to keep up with security requirements and evolving web standards. However, carrying around two complete rendering engines in a single application is rather bloaty, interaction between UI and content isn't necessarily trivial, and there may be security implications in that interaction between the two rendering engines if insecure/malicious content makes it into the UI somehow.

As a short-term damage-control measure, SeaMonkey may switch to the 45 ESR branch after 2.42, thus buying ourselves a year of support while being shielded from further removal of key features and other radical events on mozilla-release, subsequently having to figure out what to do after 45 ESR ends.

Anyway, so the discussion is going on, nothing set in stone yet other than that action must be taken and solutions figured out, but failure is not an option. 8-)

rsx11m
Moderator
 
Posts: 14425
Joined: May 3rd, 2007, 7:40 am
Location: US

Post Posted December 14th, 2015, 1:08 pm

Please stop posting about Vivaldi (or any other non-Gecko browser for that matter), this is about SeaMonkey and let's stay on topic.

patrickjdempsey

User avatar
 
Posts: 23734
Joined: October 23rd, 2008, 11:43 am
Location: Asheville NC

Post Posted December 14th, 2015, 2:41 pm

Just one note before dropping it altogether.... I've read about "customization" of the V browser. It is exactly the equivalent of cracking open the omni.ja files for Firefox or SeaMonkey and hacking them. There's nothing to be excited about when it comes to that... it's not "feature" of the browser... it's an exploitable resource. And if you don't think malware won't eventually discover that and the whole thing won't end up locked up behind a signed and encrypted archive format then you are just kidding yourself. The only reason none have bothered to do that with Firefox is because it's easier to just build an extension.

Meanwhile... I think there are positive possibilities for SeaMonkey being set free from Firefox. I know both developers and users would be happy to be free from the rapid release schedule. Good old big update once-a-year style development could come back. Also, if parts of SM have to be rewritten to handle whatever change happens, that could be really good. The themes could use a major refresh... with the exception of a few images here and there, the themes have pretty much been frozen since 2005. Modern especially suffers from serious bit rot and in some places still uses techniques from Netscape days.

I got interested in picking up that work but got dissuaded and then distracted. But here lately I've been thinking about giving it a go again. I'm talking about better-drawn graphics capable of handling hi-dpi screens and updated styling code. I'm talking about *fixing* weird code inherited from Firefox like the findbar and addons manager and about: pages.

Especially considering how much people here and in Tbird forums complain about changes from Firefox negatively impacting them, the opportunity to be uncoupled from the crazy train should elicit a bit more positive reaction. Not that it's going to be a smooth transition, but at least we won't be pulled off the cliff.
Tip of the day: If it has "toolbar" in the name, it's crap.
What my avatar is about: https://addons.mozilla.org/en-US/seamonkey/addon/sea-fox/

rn10950
 
Posts: 18
Joined: December 9th, 2014, 6:48 am

Post Posted December 14th, 2015, 3:58 pm

After I've had some time to think about it, maybe switching to ESR would be the best choice, XUL removal or not. It will allow us to stabilize the SeaMonkey UI and fix many of the UI bugs that Firefox introduces. We can base SeaMonkey off of ESR, and wait a month or so after ESR drops to release so we can fix all of the bugs introduced. That way, we can keep the UI stable while also getting security updates, and it will give us some breathing room to fix some old bugs or work on releng.

Lemon Juice
 
Posts: 777
Joined: June 1st, 2006, 9:41 am

Post Posted December 14th, 2015, 4:25 pm

From reading the chat log of a recent SeaMonkey status meeting the devs said that embedding servo is not an option. I wonder what the reasons might be - from reading about Servo in Wikipedia:
Servo provides a consistent API for hosting the engine within other software. It is designed to be compatible with Chromium Embedded Framework, an API used by Adobe and Valve Corporation to incorporate the Blink rendering engine within their own products. Allowing Servo to be dropped in as a replacement engine simplifies real-world testing.

Doesn't this mean that Servo is made to be embeddable?
*** SeaMonkey — weird name, sane interface, modern bowels ***
Mouse Gestures for SeaMonkey/Firefox
Convert Fx and TB extensions to SeaMonkey

rsx11m
Moderator
 
Posts: 14425
Joined: May 3rd, 2007, 7:40 am
Location: US

Post Posted December 14th, 2015, 4:32 pm

I see something similar in https://github.com/servo/servo/wiki/Design stating "Servo is explicitly not aiming to create a full Web browser (except for demonstration and experimentation purposes). Rather it is focused on creating a solid, embeddable engine." Thus, it would appear that #3 is an option, though still coming with the overhead of carrying two full rendering engines around. Also, the question how to interact with the embedding application remains open, and how well it will work within a Gecko-style program.

patrickjdempsey

User avatar
 
Posts: 23734
Joined: October 23rd, 2008, 11:43 am
Location: Asheville NC

Post Posted December 14th, 2015, 4:40 pm

Hmmm... I don't know enough about how Composer and Mail/News work... but I *think* the "content" areas of those are "browser" elements... so they would have to be completely rewritten in HTML to work with Servo. Also other in-content interfaces like the Data Manager would have to be rewritten... and the communication issues between segregated content and chrome is a big reason it's taken 6 years of development and e10s still isn't ready.
Tip of the day: If it has "toolbar" in the name, it's crap.
What my avatar is about: https://addons.mozilla.org/en-US/seamonkey/addon/sea-fox/

Frank Lion

User avatar
 
Posts: 20059
Joined: April 23rd, 2004, 6:59 pm
Location: ... The Exorcist....United Kingdom

Post Posted December 14th, 2015, 5:44 pm

There is opportunity here for the SM Council. Opportunity to re-appraise and re-evaluate fundamental questions and the first is - 'What is an Internet browser?' Like most things, that is not as obvious as it may seem. 'Would it be possible to just place an entire browser into an already existing sandbox? ..is another one that bears analysis.

For example, how come no one pulled me up on the following and never would have? -

Frank Lion wrote: Don't forget that the extensions and themes to turn SM into Firefox as it was years back, already exist and add less that 500kb to the total.


It is a fundamentally flawed statement. It is obviously impossible for SM to instantly backport its entire under the hood code engine back to what it was 3 or 4 years back with just a penalty of 500kb. It is nonsense...or is it? I ask because users don't think like devs and if it walks like a duck, talks like a duck, then it is a duck or, in this case, Firefox 3.6. So again, what actually is a browser?

I had this sort of re-appraisal to do on Complete Themes almost exactly 2 years ago to the day. Every single Complete Theme in the last 13 years had been using the theme replacement method, first coded up by Joe Hewitt back in 2002 (?) Vaguely reasonable method up until 4 or 5 years back and then it got harder and harder to create themes and maintain themes using that method, to the point of near impossibility.. How could there be any other possible way to make themes other than what 500+ past and present theme writers had already done?

The result of that re-assessment was a method for a Complete Theme to overlay the existing default theme. It takes 10% of the previous time to create, 10% of the previous time to maintain and flexes and adjusts to new default features being added to the core program. Pat here also played a part in the creation of my new theme template.

Of course, now, there's bound to be 'Oh yes, Frank, we always knew about all that...' Yeah? So you deliberately chose to spend 600 hours a year, instead of just 60, in maintaining your themes, did you?

So, yeah, sometimes it pays to just stop....and think 'What am I actually trying to achieve here?'
Metal Lion latest SeaMonkey & Thunderbird Themes - Sea Monkey and Silver Sea Monkey
"The only thing necessary for the triumph of evil, is for good men to do nothing." - Edmund Burke (attrib.)

tonymec

User avatar
 
Posts: 717
Joined: October 15th, 2004, 2:58 am
Location: Ixelles (Brussels Capital Region, Belgium)

Post Posted December 14th, 2015, 6:42 pm

Frank Lion wrote:I had this sort of re-appraisal to do on Complete Themes almost exactly 2 years ago to the day. Every single Complete Theme in the last 13 years had been using the theme replacement method, first coded up by Joe Hewitt back in 2002 (?) Vaguely reasonable method up until 4 or 5 years back and then it got harder and harder to create themes and maintain themes using that method, to the point of near impossibility.. How could there be any other possible way to make themes other than what 500+ past and present theme writers had already done?

The result of that re-assessment was a method for a Complete Theme to overlay the existing default theme. It takes 10% of the previous time to create, 10% of the previous time to maintain and flexes and adjusts to new default features being added to the core program. Pat here also played a part in the creation of my new theme template.

Talking of that, how many Complete Themes are still available for SeaMonkey? At https://addons.mozilla.org/en-us/seamon ... rt=updated there are exactly 27 of them, which IMHO seems damn few, and of these 27, there are 10 which were not updated even once in 2015, and of these ten, there is one whose latest update was in 2007. Of course, those which are not abandoned yet are probably the best ones. OTOH, one possible reason why there are so few of them (and also, if the user counts are to be believed, why they have so few users) is that it's tricky to get at them on AMO: From the "Themes" page (which is actually about lightweight themes) hover the mouse over the "Themes" breadcrumb near the top of the page, and a rolldown widget opens with at the very bottom "Want more customization? Try Complete Themes". Not what I would call very discoverable: apparently the idea was to kill complete themes by attrition ever since lightweight themes came around, at the time under the name "Persona": remember that «no one uses it, and anyway it isn't discoverable» is one of the excuses they use when they've decided to kill some feature. So they make it almost undiscoverable, no new users discover it, and Presto! Now you see it, now you don't.
Best regards,
Tony

barbaz
 
Posts: 1680
Joined: October 1st, 2014, 3:25 pm

Post Posted December 14th, 2015, 6:58 pm

rsx11m wrote:Another option suggested was to run SeaMonkey itself with a possibly stale Gecko version while somehow embedding a current rendering engine (and possibly Servo once ready) to keep up with security requirements and evolving web standards. However, carrying around two complete rendering engines in a single application is rather bloaty, interaction between UI and content isn't necessarily trivial, and there may be security implications in that interaction between the two rendering engines if insecure/malicious content makes it into the UI somehow.

+1 to this idea

Frank Lion wrote: 'What is an Internet browser?'

In my mind it's an application that allows the user to view Web pages the way that user wants to view them. Meaning, if the user wants to, say, do any of: disable JS and other active content, inject custom buttons/links into the page for source form, inspect the DOM and/or run custom JS snippets, etc... they can do it, either by built-in application functionality or installing extensions. It's also important to have control over how the browser requests pages and how the browser presents itself to a website, in order to e.g. spoof browser profile information for privacy, add/alter request headers, or effectively disable certain JS properties on a per-site basis with early-running JavaScript.

A web browser is NOT something that has any specific site or service integrated into it so deeply that it cannot be removed (I'm saying no social-network integration and nothing like Pocket or Hello), this stuff is better left as extensions (or separate apps entirely) so that the user can 100% choose whether to have it or not.

Basically what I'm saying is that the level of control you can get with SeaMonkey now is IMO critical to have for a good, safe Web browsing experience; and that a "proper" Internet browser will always provide the ability for the user to take full control over all of said browser's interactions with the Internet.

Just my 2 cents.
Last edited by barbaz on December 14th, 2015, 7:02 pm, edited 1 time in total.
*Always* check the changelogs BEFORE updating that important software!

rsx11m
Moderator
 
Posts: 14425
Joined: May 3rd, 2007, 7:40 am
Location: US

Post Posted December 14th, 2015, 7:02 pm

...what is a web browser? Whatever it is, SeaMonkey is definitely more. ;-)
Let's keep in mind that we also have to keep the Mail/News side happy.

Also note that all of those ideas are great as long as someone is also willing to commit time and effort into implementing it, and that may be stuff that the current limited number of active developers and volunteers aren't readily familiar with.

And on a side note, today was merge day and 45 is now on comm-aurora, which would be the future 45 ESR if that's the way to go (i.e., either accept the string freeze or convince localizers that an occasional limited-impact patch with string changes will be accepted on the comm-esr45 branch).

barbaz
 
Posts: 1680
Joined: October 1st, 2014, 3:25 pm

Post Posted December 14th, 2015, 7:09 pm

Well, my definition of Internet browser can apply also to email client - in short, something that works with emails as the user wants it to. Important user-control features there IMO include request blocking (block ALL requests originating from emails! Client-internal only please), ability to alter mail headers, ability to display emails in plain text or however the user wants, ability to save emails, choosing when & how to display or save attachments, the junk filter and manual filters, "Empty junk" button, choosing which folder is which (like, this folder is set as Trash and that one is set as Junk even though they're not "defaults" for that), and ability to tweak "advanced" parameters to work around quirks of specific mail servers.

Can't speak for News as I don't use that.
*Always* check the changelogs BEFORE updating that important software!

Return to SeaMonkey General


Who is online

Users browsing this forum: No registered users and 2 guests