5 ms·

>Trails is built and maintained by former members of the Sails.js core team Just no. I was working with Sails.js for about a year now on production systems - i

>Trails is built and maintained by former members of the Sails.js core team Just no. I was working with Sails.js for about a year now on production systems - it's horrible. Without a doubt I can say that it's the worst framework I've ever worked with. Especially their so called "orm" - we actually had to switch it off and work with pure mysql queries instead because it can't even do migrations (it deletes everything from db, stores in ram and re-insert). Seriously, don't believe the hype, I bet this will be just as bad as Sails.

roblabla 11y ago
While I haven't played with Sails.JS, I've heard many people complain about how bad it was. However, just because the people building this new framework used to work on the Sails.JS Framework, you're going to dismiss it as being "automatically bad" ? I didn't look at Trails.JS so I don't know how it looks, but this kind of attitude seems extremely toxic and detrimental to the Open Source space. It makes me sad this is the top comment. If anything, this gives them the chance to start anew, equipped with the knowledge of past mistakes, to avoid making them again.
- dhruvbhatia 11y ago
  In his defence, he is taking a stance equipped with the knowledge of his own personal past mistakes, to avoid making them again.
  - paulddraper 11y ago
    And his past mistake was using Sails. If the implementation of the Sails as a Rails-like JS framework was as bad as he says, be shouldn't use it for his next project. A bunch of members of the community apparently had similar thoughts. As long as the idea of Rails-like JS framework isn't fundamentally flawed, they seem to be doing the logical next step: make something better.
- im_down_w_otp 11y ago
  Can I dismiss it because it's like the 50th "modern MVC framework" to be made for Node.js? Are all the other ones so hopelessly broken that it required making an entirely new one?
  - smokeyj 11y ago
    Dismiss it, but you don't have to tell the entire world about your uninformed opinion. It's not as valuable as you suspect.
    - im_down_w_otp 11y ago
      Opinion? Both sentences were questions. The first included a hyperbolic approximation (i.e. "50th") and the second included a legitimate direct question that infers an indirect question about the weird pace of churn and factionalization that seems to infect the Node.js community more than any other community. If every other month a new thing was made, marketed, HN'd, and Twitter-culted that did basically the same thing as Nginx but had a slightly different config file format or shared-object handling mechanism or some other random detail I'd think that was also weird and somewhat superfluous. It makes me wonder if building MVC frameworks has become so technically rote (time investment notwithstanding) that the simplicity of it has pushed them into Parkinson's Law of Triviality territory, such that everybody has their two-cents on how this weird metaprogramming trick is better than this other weird metaprogramming trick so now I can write my routes and handlers in this shape of quasi-DSL instead of that other quasi-DSL, and so then goes and makes a whole "new"... sorry... "modern" MVC framework. How far away are we from realistically being able to replace the idiom, "bike-shedding" with, "MVC-frameworking"?
      - smokeyj 11y ago
        I'm just saying if you have a criticism share it. If your problem is being an nth iteration then that's not really comment worthy. Sure the node ecosystem has an abundance of modules and frameworks. I don't consider it harmful and it seems the developer mindshare sorts through these "problems" with ease.
- bnb 11y ago
  100% agree with you. I don't agree with the comment you replied to, but it's _not_ a good sign that they stole another project's logo. https://tent.io/ https://tent.io/
  - yejodido 11y ago
    Hi bnb, it hurts to see the word "stole" and judgement passed so hastily, so I thought I'd take a quick second to reply. It was only just this morning with the attention from Hacker News that we learned about the similarities between the two logos/brandmarks, and we're actively coming up with something new and different. We'll keep the community posted on the status of the new designs here: https://github.com/trailsjs/trails/issues/47 https://github.com/trailsjs/trails/issues/47 Being open, compassionate, and transparent is a huge part of what we're hoping to do with Trails, so chatting with us directly is always an option. We hope to see you on gitter! All best :)
  - traviswebb 11y ago
    We've been in contact with the owner of tent, and are responding appropriately. Our logo unfortunately does bear a strong resemblance to theirs, which our designers are working to resolve. I'd hitherto not heard of tent nor seen their logo, and neither had anyone else on our team. Saying that we "stole" it is fine for sloppy internet-talk, but it imbues your statement with a level of accusatory malice that we should try to avoid.
- mikermcneil 11y ago
  Travis did not play a meaningful role in the development of Sails. He was a contributor for less than a year and made a total of 37 commits: https://github.com/balderdashy/sails/commits?author=tjwebb https://github.com/balderdashy/sails/commits?author=tjwebb That said, I wish Trails the best-- I just think it's important to point out information that is being misrepresented. If you have any technical issues with Sails, I would appreciate hearing about those on GitHub.
kevinob11 11y ago
I think it is only part of the core team, frustrated by the lack of progress on sails, so maybe they want some of the same changes you do. Also, while I had my own frustrations with waterline, the auto migrations were meant to speed up development, they were never meant to be used in production and can be turned off with a single line if code.
- romanovcode 11y ago
  Once you start shipping to production you care far less about "speed of development" and "prototyping things" and more about "does this actually work". Don't get me wrong, Sails might be very good for prototyping something, but I could recommend it only if your only goal was to prototype something.
  - kevinob11 11y ago
    Fair enough, though I do have some experience with shipping to production. I do have quite a few other issues with sails and waterline (ORM). Waterline provided lifecycle hooks, but limited their power and had poor documentation on extending them. Same for sails core itself, it had hooks for writing modules and extensions, but no one outside of the core team knew well how to use them. As for migrations I ended up doing all of my own migrations with knex, I'd love to see knex or sequelize be the core of a new ORM, rather than a quickly cobbled together sql module. Trails has promised something better on a lot of these fronts through modularity, I hope they deliver.
    - clay_to_n 11y ago
      Lack of documentation has probably been my biggest gripe with Sails. The documentation on their website was broken for almost a year ("but it's on github! Go find it there"). They had a Getting Started "Sailscasts" series that was helpful, and then they released the beginning of the V2 videos (using a SPA w/ Angular), but only got three videos in.
      - sgress454 11y ago
_____s 11y ago
I'd give it a chance. They may have made poor decisions earlier, but there's a good chance they know that, too, and they would have learned from it. A newer framework from the same people has at least one thing going for them: they aren't likely to make the same mistakes.
paulddraper 11y ago
It sounds like you are in violent agreement with the Trails team, as they share your opinion. Hence, Trails.
dimgl 11y ago
I try to stay away from big frameworks that try to do this very thing because development and production experiences are two completely different things. I've been through nightmares scenarios where things don't work correctly or new technical debt is introduced after deployment and smaller frameworks usually alleviate that pain point for me.
mikermcneil 11y ago
Sorry to hear you had a bad experience. The auto-migration support in Sails is a development-only feature, but for production, there is a great project from Danni Friedland called sails-migrations: https://github.com/BlueHotDog/sails-migrations https://github.com/BlueHotDog/sails-migrations You might also want to check out sails-db-migrate: https://github.com/building5/sails-db-migrate https://github.com/building5/sails-db-migrate At Treeline, we're using sails-migrations in production.

roblabla 11y ago

While I haven't played with Sails.JS, I've heard many people complain about how bad it was. However, just because the people building this new framework used to work on the Sails.JS Framework, you're going to dismiss it as being "automatically bad" ? I didn't look at Trails.JS so I don't know how it looks, but this kind of attitude seems extremely toxic and detrimental to the Open Source space. It makes me sad this is the top comment. If anything, this gives them the chance to start anew, equipped with the knowledge of past mistakes, to avoid making them again.

dhruvbhatia 11y ago
In his defence, he is taking a stance equipped with the knowledge of his own personal past mistakes, to avoid making them again.
- paulddraper 11y ago
  And his past mistake was using Sails. If the implementation of the Sails as a Rails-like JS framework was as bad as he says, be shouldn't use it for his next project. A bunch of members of the community apparently had similar thoughts. As long as the idea of Rails-like JS framework isn't fundamentally flawed, they seem to be doing the logical next step: make something better.
im_down_w_otp 11y ago
Can I dismiss it because it's like the 50th "modern MVC framework" to be made for Node.js? Are all the other ones so hopelessly broken that it required making an entirely new one?
- smokeyj 11y ago
  Dismiss it, but you don't have to tell the entire world about your uninformed opinion. It's not as valuable as you suspect.
  - im_down_w_otp 11y ago
    Opinion? Both sentences were questions. The first included a hyperbolic approximation (i.e. "50th") and the second included a legitimate direct question that infers an indirect question about the weird pace of churn and factionalization that seems to infect the Node.js community more than any other community. If every other month a new thing was made, marketed, HN'd, and Twitter-culted that did basically the same thing as Nginx but had a slightly different config file format or shared-object handling mechanism or some other random detail I'd think that was also weird and somewhat superfluous. It makes me wonder if building MVC frameworks has become so technically rote (time investment notwithstanding) that the simplicity of it has pushed them into Parkinson's Law of Triviality territory, such that everybody has their two-cents on how this weird metaprogramming trick is better than this other weird metaprogramming trick so now I can write my routes and handlers in this shape of quasi-DSL instead of that other quasi-DSL, and so then goes and makes a whole "new"... sorry... "modern" MVC framework. How far away are we from realistically being able to replace the idiom, "bike-shedding" with, "MVC-frameworking"?
    - smokeyj 11y ago
      I'm just saying if you have a criticism share it. If your problem is being an nth iteration then that's not really comment worthy. Sure the node ecosystem has an abundance of modules and frameworks. I don't consider it harmful and it seems the developer mindshare sorts through these "problems" with ease.
bnb 11y ago
100% agree with you. I don't agree with the comment you replied to, but it's _not_ a good sign that they stole another project's logo. https://tent.io/ https://tent.io/
- yejodido 11y ago
  Hi bnb, it hurts to see the word "stole" and judgement passed so hastily, so I thought I'd take a quick second to reply. It was only just this morning with the attention from Hacker News that we learned about the similarities between the two logos/brandmarks, and we're actively coming up with something new and different. We'll keep the community posted on the status of the new designs here: https://github.com/trailsjs/trails/issues/47 https://github.com/trailsjs/trails/issues/47 Being open, compassionate, and transparent is a huge part of what we're hoping to do with Trails, so chatting with us directly is always an option. We hope to see you on gitter! All best :)
- traviswebb 11y ago
  We've been in contact with the owner of tent, and are responding appropriately. Our logo unfortunately does bear a strong resemblance to theirs, which our designers are working to resolve. I'd hitherto not heard of tent nor seen their logo, and neither had anyone else on our team. Saying that we "stole" it is fine for sloppy internet-talk, but it imbues your statement with a level of accusatory malice that we should try to avoid.
mikermcneil 11y ago
Travis did not play a meaningful role in the development of Sails. He was a contributor for less than a year and made a total of 37 commits: https://github.com/balderdashy/sails/commits?author=tjwebb https://github.com/balderdashy/sails/commits?author=tjwebb That said, I wish Trails the best-- I just think it's important to point out information that is being misrepresented. If you have any technical issues with Sails, I would appreciate hearing about those on GitHub.

kevinob11 11y ago

I think it is only part of the core team, frustrated by the lack of progress on sails, so maybe they want some of the same changes you do. Also, while I had my own frustrations with waterline, the auto migrations were meant to speed up development, they were never meant to be used in production and can be turned off with a single line if code.

romanovcode 11y ago
Once you start shipping to production you care far less about "speed of development" and "prototyping things" and more about "does this actually work". Don't get me wrong, Sails might be very good for prototyping something, but I could recommend it only if your only goal was to prototype something.
- kevinob11 11y ago
  Fair enough, though I do have some experience with shipping to production. I do have quite a few other issues with sails and waterline (ORM). Waterline provided lifecycle hooks, but limited their power and had poor documentation on extending them. Same for sails core itself, it had hooks for writing modules and extensions, but no one outside of the core team knew well how to use them. As for migrations I ended up doing all of my own migrations with knex, I'd love to see knex or sequelize be the core of a new ORM, rather than a quickly cobbled together sql module. Trails has promised something better on a lot of these fronts through modularity, I hope they deliver.
  - clay_to_n 11y ago
    Lack of documentation has probably been my biggest gripe with Sails. The documentation on their website was broken for almost a year ("but it's on github! Go find it there"). They had a Getting Started "Sailscasts" series that was helpful, and then they released the beginning of the V2 videos (using a SPA w/ Angular), but only got three videos in.
    - sgress454 11y ago

_____s 11y ago

I'd give it a chance. They may have made poor decisions earlier, but there's a good chance they know that, too, and they would have learned from it. A newer framework from the same people has at least one thing going for them: they aren't likely to make the same mistakes.

paulddraper 11y ago

It sounds like you are in violent agreement with the Trails team, as they share your opinion. Hence, Trails.

dimgl 11y ago

I try to stay away from big frameworks that try to do this very thing because development and production experiences are two completely different things. I've been through nightmares scenarios where things don't work correctly or new technical debt is introduced after deployment and smaller frameworks usually alleviate that pain point for me.

mikermcneil 11y ago

Sorry to hear you had a bad experience. The auto-migration support in Sails is a development-only feature, but for production, there is a great project from Danni Friedland called sails-migrations: https://github.com/BlueHotDog/sails-migrations https://github.com/BlueHotDog/sails-migrations You might also want to check out sails-db-migrate: https://github.com/building5/sails-db-migrate https://github.com/building5/sails-db-migrate At Treeline, we're using sails-migrations in production.